The Nuclear Fuel Cycle

Transcription

CHAPTER 21
Contents
21.1.
21.2.
21.3.
21.4.
21.5.
21.6.
21.7.
21.8.
21.9.
21.10.
21.11.
21.12.
21.13.
21.14.
21.15.
21.16.
Production of fuel elements
Power generation
Composition and properties of spent fuel elements
21.3.1.
Fission products
21.3.2.
Actinides
21.3.3.
Decay heat and physical properties
Management of spent fuel
21.4.1.
Transport of spent reactor fuel
21.4.2
Interim storage facilities
Alternative fuel cycles
21.5.1.
The uranium once-through (UOT) option
21.5.2.
The uranium - plutonium (U-Pu) fuel cycle
21.5.3.
The thorium - uranium (Th-U) fuel cycle
Reprocessing of uranium and mixed oxide fuels
21.6.1.
Head end plant
21.6.2.
Separation methods
21.6.3.
Purex separation scheme
21.6.4.
Engineering aspects and operation safety
Reprocessing of thorium fuels
Wastes streams from reprocessing
21.8.1.
Gaseous wastes
21.8.2.
Liquid wastes
21.8.3.
Organic wastes
21.8.4.
Solid wastes
21.8.5.
Environmental releases from reprocessing plants
Treatment of and deposition of low and medium level wastes
Tank storage of high level liquid wastes
Options for final treatment of high level waste
21.11.1. Dispersion into sea and air
21.11.2. Partitioning
21.11.3. Disposal into space
21.11.4. Remote deposition
21.11.5. Transmutation
Solidification of high level liquid wastes
Deposition in geologic formations
21.13.1. Properties of geologic formations
21.13.2. Waste conditioning before final storage
21.13.3. Repository projects
Beneficial utilization of nuclear wastes
Exercises
Literature
583
585
588
592
593
596
599
599
599
600
601
601
602
604
605
605
608
611
613
615
616
616
618
619
619
620
620
622
623
624
626
628
629
630
631
633
633
636
637
639
640
641
584
Radiochemistry and Nuclear Chemistry
FIG. 21.1. Fuel cycle alternatives. Annual flows of materials in a 10 GWe LWR UOT program
are indicated.
The nuclear fuel cycle comprises the handling of all fissile and fertile material necessary for
nuclear power production and of the radioactive products formed in this process (Fig. 21.1).
The fuel cycle is suitably divided into a front end and a back end part, where the nuclear power
station is the dividing line. The front end comprises uranium exploration,
585
mining, and refining (§5.5), isotope enrichment (§2.8), and fuel element fabrication (§21.1).
Reactor operation involves fuel behavior during operation, canning corrosion etc., while the
back end involves reprocessing and radioactive waste ("radwaste") handling. Health and
environmental aspects are important in all these steps, but being of a more general nature (see
Ch. 18 and 22), they are not considered as "steps" in the nuclear fuel cycle.
The nuclear fuel used in almost all commercial reactors is based on uranium in the form of
UO2 , either enriched so that the 235U content has been increased to a few percent, or ! less
commonly ! with the natural 0.7% abundance of the fissile isotope 235U. Some power reactors
also use fuel containing depleted uranium (~0.3% 235U) in which plutonium is bred and/or Pu
mixed with 238U as a replacement for 235U ("mixed oxide fuel"). 232Th, in which fissile 233U
is bred, has also been used in a few cases.
Whether based on uranium, thorium, or plutonium, a fuel must be capable of resisting
temperatures considerably above 1000EC without physical or chemical deterioration due to heat
or to radiation. Metallic fuels have the high heat conduction necessary to minimize temperature
gradients. Uranium melts at 1130EC and plutonium at 640EC. Moreover, metallic uranium has
three and plutonium six allotropic forms between room temperature and their melting points.
As a consequence, either the separate or combined effects of the radiation field, the high
pressure during operation, and the high temperature can cause recrystallization into different
allotropic forms with significantly different volume. Volume changes within the fuel element
during operation cause mechanical deformations, reduce the mechanical strength, and increase
the problem of corrosion even if the elements are clad in another corrosion resistent metal.
With the exception of some older gas cooled reactors, power reactors use ceramic pellets of
UO2, PuO2, and ThO2, or a mixture of these oxides, as fuels. UC has also been tested in some
reactors. The size of the cylindrical pellets is ~1 × 1 cm (diameter × height). Fuel rods
consisting of ceramic fuel pellets stacked in metallic tubes of zircaloy or stainless steel are quite
temperature resistant, do not have the phase transformations of the metals, and have greater
resistance to radiation effects. Unfortunately, the heat conduction is not as good as in the
metallic fuel elements, and as a result rather high temperature gradients (up to 300EC mm!1)
often exist in the ceramic elements.
21.1. Production of fuel elements
The normal raw material for production of UO2 based fuels is enriched uranium in the form
of UF6 , which is delivered in special containers. By heating the container to ~100EC it is
possible to transfer the hexafluoride as gas to the conversion plant, where UO2 powder is
produced, see Fig. 21.2. Several possible reactions can be used, e.g. hydrolysis of UF6 by
dissolution in water
UF6(g) + 2H2O 6 UO2F2(aq) + 4HF(aq)
followed by precipitation with ammonia
2UO2F2(aq) + 6NH4OH(aq) 6 (NH4)2U2O7(s) + 4NH4F(aq) + 3H2O
586
FIG. 21.2. Production of UO2-fuel rods.
after filtration, washing and drying the diuranate is converted to UO2 by reduction with
hydrogen at 820EC.
Another often used reaction sequence begins with the formation and precipitation of
ammonium uranyl carbonate (AUC) by reaction of UF6 with water, CO2 and NH3, see Figure
21.2. The AUC is reduced to UO2 by reduction in a fluidized bed using a mixture of hydrogen,
nitrogen and steam and cooled to room temperature in a mixture of air and nitrogen.
The UO2 powder is pressed into "green" pellets of slightly larger dimensions than the final
product. These have ~50% of the theoretical density. The green pellets are sintered at
~1700EC in a dry H2 atmosphere which gives a small controlled oxygen excess in the product;
UO2+x, where x ~0.05 is normally desired for best fuel performance. As exact dimensions of
the pellets are needed in order to fit into the cladding tubes, the sintered pellets are ground to
final shape.
Typically a density of ~10 400 kg m!3 is desired, which corresponds to a porosity of ~5%.
If the density is too high the pellets swell excessively during irradiation due to the volume of
the fission products gaseous at the operating temperature. Strongly swelling pellets may cause
deformation, and failure, of the can. On the other hand a too low density can cause an initial
shrinking of the pellet leading to an increased pellet ! can gap, thereby reducing the heat
transfer coefficient and increasing pellet temperature.
The finished pellets are put into a zircaloy tube with a welded bottom plug. Thereafter the
pellet column hold-down spring and top plug are fitted and the resulting fuel pin sealed by
electron beam welding, see Figures 19.12 and 21.2.
In order to improve heat transfer between pellets and can and to reduce pressure induced strain
during operation, the space between pellets and tube is usually first evacuated and then
pressurized with helium through a hole in the top end plug. It important to remove all traces of
water as remaining water may form hydrogen by reaction with hot UO2 during operation. The
hydrogen reacts with the inside of the can forming zirconium hydride which may cause can
failure.
587
Heat conduction between fuel and can is further improved by using bonding materials such
as molten sodium, graphite powder, etc. The bonding material should also provide some
lubrication between pellets and can. The canning material itself must not only be corrosion
resistant to the coolant at all temperatures but should react with neither the fuel nor the bonding
material.
The can should be as thin as possible, consistent with satisfactory mechanical strength and
corrosion resistance (Fig. 19.12(a)). To reduce the danger of hydride formation a protective
oxide layer can be produced by autoclaving the tube before filling it with pellets. In case of UO2
pellets in zircaloy, the bonding material, e.g. graphite, is put onto the inner surface of the
zircaloy tubes before the pellets are introduced. In case of stainless steel clad fast reactor fuel
the production and assembly is similar, but the bonding is usually by sodium metal.
The purpose of cladding the fuel is to protect it against corrosion and to protect the coolant
from radioactive contamination by the fission products from the fuel element. Aluminum has
been used in water-cooled reactors, but at temperatures > 300EC zirconium alloys show
superior strength. At steam temperatures > 400EC zirconium absorbs hydrogen, which
increases brittleness, so stainless steel becomes preferable. In sodium cooled fast reactors,
stainless steel is normally used. The most common alloys are zircaloy-2 (containing 1.58 Sn
and 0.3% Cr, Ni, and Fe) and stainless steel type 302B (containing 10% Cr and 8% Ni).
Stainless steel is not used at lower temperatures because of its larger neutron capture
cross-section: F 0.23 b for Al, 0.18 b for zircaloy-2, and about 3 b for 302B steel.
Metallic uranium is usually produced by conversion of UF6 to UF4 followed by metallothermic
reduction of UF4 by magnesium or calcium metal; however, several other methods exist.
Metallic fuel is encased in a canning (cladding) of aluminum, magnesium, or their alloys. Fuel
for high flux research reactors based on highly enriched uranium (>10% 235U) is often made
in the form of uranium metal alloys (or compounds like USi2) canned in aluminum to improve
mechanical and thermal stability.
The fuel elements for use in high temperature gas cooled reactors consist of graphite rods or
balls filled with oxide or carbide kernels produced by the sol-gel process. The kernels are
covered by several layers of graphite and silicon carbide achieved by pyrolyzing methane or
acetylene in a fluidized bed of the kernels.
Fuel cost and performance is an important part of the economy of power reactors.
Approximately 20% of the expense of the electrical production in a power reactor can be
attributed to the cost of the fuel. This is due about equally to the expense of the consumption
of fissile material and to the production and, when applicable, reprocessing costs or
intermediate storage costs. In the fast breeder reactors, it is anticipated that fuel costs would
be substantially lower because of a higher burn-up.
When mixed uranium-plutonium oxide (MOX) fuel elements are used as, for example, in
plutonium recycling (< 5% PuO2) in LWRs or in fast breeders (# 15% PuO2), the UO2!PuO2
mixture must be very intimate. This can be achieved by coprecipitation of the tetravalent
actinides, normally as oxalates, followed by calcination. However, in industrial scale production
of MOX fuel a rich mixture containing 15!20% PuO2 is first very finely ground and then
diluted with coarse grained pure UO2. The final mixture is pressed into pellets and sintered
similarly to UO2-fuel pellets. This yields a fuel which can be dissolved to 99% (or better) in
nitric acid. MOX fuel elements have been regularly added to the cores
588
of many European LWRs for many years without any technical difficulties. Mixed uranium and
thorium oxide fuels have also been used in a few heavy water reactors, see §20.1.1.
Instead of ground powders, spherical fuel particles can be used as a starting material. This has
advantages with respect to fabrication, reactor utilization, and fuel reprocessing. These oxide
or carbide particles are very small, < 1 mm in diameter. The particles are produced by the
sol!gel process, which in principle consists of the following steps:
(i)
An aqueous colloidal solution or the actinide or actinide mixture is prepared. The
actinide(s) may be in the form of a hydrated complex of high concentration (3!4 M).
(ii)
The solution is added to an inert solvent, which dehydrates the complex and causes the
droplet to gelate. In one technique, hexamethylenetetramine, (CH2)6N4, is added to
the aqueous solution, which is added dropwise to a hot (~95EC) organic solvent. The
heat causes the amine to decompose, forming NH3, which leads to hydroxide
precipitation in the droplet. The droplet dehydrates and solidifies rapidly, forming a
"kernel".
(iii)
The kernels are washed, air dried at 150-200EC, and - in the case of uranium reduced by hydrogen gas at higher temperature to form UO2.
(iv)
The kernels are sintered at high temperature in an inert atmosphere.
Kernels of actinide carbides can be made in a similar manner. In many cases the kernels are
covered by the addition of protective layers of graphite, silica or silicon carbide. The kernels
are placed in fuel rod cans, pressed into pellets, or incorporated in a graphite matrix for use in
high temperature reactors.
21.2. Power generation
During operation a large amount of heat is generated inside the fuel and has to be transferred
to the coolant. Assuming a constant power per unit volume in the fuel, and fuel pins so long that
longitudinal conduction can be neglected, the temperature profile can be estimated from the
specific power (p, W m!3) in the fuel from the following equations using the notation for radii
from Figure 21.3.
Inside the fuel
T(r) = T(rf) + p (rf2 ! r 2) (4 kf)!1
(21.1)
where T(r) (EC) is the temperature at radius r (m) inside the fuel pellet, T(rf) (EC) the
temperature at the pellet surface, rf the pellet outer radius (m) and kf the heat conductance in
the fuel (W m!1 K!1).
Across the gap between fuel and can
)Tgap = p rf (ri - rf) (2 kg)!1 = p rf (2 "g)!1
(21.2)
where )Tgap is the temperature difference across the fuel - can gap (EC), ri the inner radius of
the can (m), kg the heat conductivity across the gap (W m!1 K!1) and " g is the heat transfer
coefficient across the gap (W m!2 K!1).
For a thin can
589
FIG. 21.3. Calculated temperatures in a BWR UO2-fuel pin during operation at high load.
)Tcan = p rf2 (rc - ri) (2 ri kc)!1
(21.3)
where )Tcan is the temperature difference across the can (EC), rc the outer radius of the can (m)
and kc the heat conductivity in the can (W m!1 K!1).
Between the can surface and coolant,
)To = p rf2 (2 rc h)!1
(21.4)
where )To is the temperature difference between the outer surface of the can and the coolant
(EC) and h the film coefficient for heat transfer between can and coolant (W m!2 K!1). The
film coefficient is affected by the coolant velocity along the surface, the temperature gradient
and also by the onset of boiling in a BWR.
Temperatures are best calculated by starting with eqn. (21.4) and proceeding inward to the
center. A typical temperature profile is shown in Figure 21.3. However, in a more accurate
calculation we must also consider the variation in specific power with fuel radius (because of
self-screening effects etc.), the variation of heat conductivity (and film coefficient) with
temperature, the change in heat conductivity of the fuel caused by accumulation of fission
products, by pellet breakup, and by the densification of the fuel caused by high operating
temperatures; the heat conductivity of UO2 as function of temperature is shown in Figure 21.4
and some typical data at room temperature are given in Table 21.1.
As a result of these heat gradients it has been found that ceramic fuel elements may melt in
the center (2865EC mp for pure UO2) at high loading even though the surface temperature is
much below the melting point. High center temperatures, especially in
590
FIG. 21.4. The heat conductivity of UO2 as function of temperature.
breeder fuel, may cause so much densification that a central hole is formed during operation.
The steep temperature gradients in the fuel during operation can cause a gradient of thermal
expansion in the pellet; the expansion increasing from the surface to the center. The
TABLE 21.1. Data (25EC) on some materials used in making nuclear fuel
Material
Density
(kg m!3)
Melting
point
(K)
Thermal
conductivity
(W m!1 K!1)
Heat
capacitivity
(J kg!1 K!1)
Thermal linear
expansion coeff.
(K!1)
2023
3663
1405
3138
912.5
2663
933.4
~650
1673
~2090
~2120
41
0.56
25
Fig. 21.4
8
~6
238
~167
19
118
256
116
360
137
258
903
~1 024
500
11.2×10!6
__________________________________________________________________________________________________________________
Th (metal, "-)
ThO2
U (metal, "-)
UO2
Pu (metal, "-)
PuO2
Aluminum
Magnox A12‡
SS (type 304)
Zircaloy 2†
Zircaloy 4‡‡
11 720
10 001
19 070
10 970
19 860
11 510
2 700
1 740
8 030
6 550
6 440
16
330
13.5×10!6
14×10!6
57×10!6
11.0×10!6
23.2×10!6
26×10!6
18×10!6
5.2×10!6
4.4×10!6
__________________________________________________________________________________________________________________
†
Zr + 12-17‰ Sn, 0.7-2‰ Fe, 0.5-1.5‰ Cr, 0.3-0.8‰ Ni;
Cr; ‡ 0.8% Al, 0.01% Be; cf. Magnox ZA with 0.5-0.65% Zr
‡‡
Zr + 12-17‰ Sn, 1.8-2.4‰ Fe, 0.7-1.3‰
591
FIG. 21.5. (a) Schematic drawing of fuel pellet deformation during operation and (b)
autoradiograph of a cut spent fuel pellet showing the typical pattern of cracks.
induced stresses lead to formation of a series of radial and annular cracks in the pellet and a
deformation during operation, see Figure 21.5(a). Upon cooling the cracks close, but the
characteristic pattern of cracks can be seen when a used fuel pin is cut and inspected, see Figure
21.5(b).
During operation, a slow corrosion of the can is unavoidable. As long as the corrosion
products stick to the surface, corrosion rates drop with time. For zircaloy clad fuel in water
cooled reactors the corrosion rate follows a parabolic equation (in the normal operating
temperature range)
ds/dt = (k / s) e!u
(21.5a)
where s is the thickness of the zirconium dioxide layer (m), t is the exposure time (s), k the rate
constant (3.937×10!5 m2 s!1) and u is given by
u = )E (R Tc)!1
(21.5b)
where R is the gas constant (8.318 J mole!1 K!1), )E the activation energy (1.905×105 J
mole!1) and Tc the surface temperature of the can (K). The zirconium - steam reaction becomes
very violent above ~1200EC, see §19.15. At about the same temperature a less well studied
reaction between UO2 and Zr begins in the fuel - cladding gap leading to the formation of a
metallic U+Zr melt and ZrO2.
Accumulation of the various fission products (and other impurities) occur where their chemical
potential is at minimum. In contrast to an isothermal system where concentration gradients tend
to disappear, the large temperature differences in operating nuclear fuel can result in a lower
chemical potential at a higher concentration leading to an increase in the
592
concentration gradient. Hence, some fission products, e.g. the noble gases, Cs and I, migrate
to the fuel - cladding gap (lowest temperature), whereas others, e.g. Zr and Nb, migrate to the
center line (highest temperature), cf. Fig. 21.5(b). A penetration of the can during reactor
operation thus leads to an initial rapid release of those fission products which accumulated in
the gap followed by a slow release of fission products present in the cracks and at grain
boundaries and finally by a much slower dissolution and release of fuel (U, other actinides and
other fission products).
The high concentration of uranium (and/or Pu) atoms in the fuel in combination with strong
resonance peaks at certain neutron energies also leads to self-screening effects (n-flux
depression at energies with large reaction cross sections), especially for 238U(n,(), 239Pu(n,()
and 239Pu(n,f) (see Fig. 19.3). Thus, most of the Pu formed is located near the fuel surface and
relatively little at the fuel center. This is easily seen in "-autoradiographs of spent fuel pins
where most of the ":s are found in a ring near the fuel surface. Hence, thinner fuel pins yield
a better fuel utilization than thicker pins, but thinner pins means higher fuel fabrication costs
for the same amount of uranium. In practice an economic compromise results in pin diameters
slightly less than 1 cm.
Radiation effects and oxidation causes changes in the tensile properties of the canning
material. Fresh fuel has a can that is very ductile whereas the can of spent high burn-up fuel
normally is hard and brittle.
In an operating power reactor, only part of the fuel is replaced annually; e.g. 1/5 to 1/4 of the
total number of fuel elements. The most burnt up fuel elements are removed from the core as
spent fuel and replaced by fresh fuel. In order to achieve as even as possible heat generation
in the core (permitting the highest power output), the fresh fuel elements are mostly loaded in
the outer core regions whereas partly spent fuel is moved in towards the center. This results in
a checker board pattern of fuel of varying age in the reactor.
21.3. Composition and properties of spent fuel elements
The composition of spent reactor fuels varies as a function of input composition (kinds and
amounts of fissile and fertile atoms), neutron spectrum, flux, fluency (or burnup), design of
the pins and fuel elements, positions occupied in the reactor during operation, and the cooling
time after removal from the reactor. A harder neutron spectrum increases fertile to fissile
conversion (Fig. 19.3). Hence, after refuelling, some BWRs are initially operated at the highest
possible void fraction in order to maximize conversion of 238U to plutonium. This permits a
higher final burnup of the fuel. Increased burnup increases the concentration of fission products
and larger amounts of higher actinides are formed (Figs. 16.4 and 19.7). Thinner fuel pins
increases conversion to and burning of plutonium due to less self screening. A high neutron flux
results in more high order reactions (§15.3), while a long irradiation time produces relatively
larger amounts of longlived products. With increased cooling time the fraction of shortlived
products is reduced.
Because of such effects, spent uranium fuel elements from PWR, BWR, HWR, GCR and
FBR differ in composition both from each other and between fuel batches from the same
reactor. Furthermore, the composition differs between pins in the same fuel element and for
each pin also along its length, especially when initial burnable poison concentration and
enrichment is graded along pins. The difference is not so large that very different fuel
593
TABLE 21.2. Calculated composition after 10 y cooling of 1 t U as 3.2% enriched UO2 fuel with
33 MWd/kg U burnup at a mean flux of 3.24×1018 n m!2 s!1 in a typical PWR
cycles (e.g. other reprocessing schemes) are required as long as the fuel is based on uranium
metal or uranium dioxide. In the following subsections we mainly discuss uranium dioxide fuel
elements, and, more specifically, LWR elements. Fission product and actinide yields for a
typical PWR fuel are given in Table 21.2.1 The uranium once through part of Figure 21.1
shows the annual materials flow in a mixed BWR!PWR conglomerate with a total average
power of 10 GWe (i.e. about twelve 1000 MWe plants) running at full power for 7000 hours
per year (load factor ~80%).
21.3.1. Fission products
About 34 kg fission products (FP, including gaseous) are formed in each initial ton of uranium
irradiated to 33 MWd/kg. To accommodate mixed oxide fuels (i.e. fuels initially containing
both U and Pu), burnup, composition, activities, etc, are usually normalized to the amount of
initially present heavy metal, IHM. The composition of spent fuel varies with burnup, power
history and reactor. The formation rates of the various primary fission products depend on the
fission rate, chain yields of the fissioning nuclide and on the corresponding charge distribution
(Ch. 14). Increased fission of heavier actinides displaces the lower mass peak in the yield curve
towards higher mass numbers while the heavier mass peak remains about the same (Fig. 14.9).
An increased contribution from fast fission increases the yield in the valley between the peaks.
Because of the continued n-irradiation, secondary n,(-reactions occur with the primary fission
products and their daughters (Ch.
1
As the exact composition of spent fuel varies considerably and depend on many factors, the reader will find slightly
varying figures in this text.
594
FIG. 21.6. Typical variation of the ratio between the
burnup.
134Cs
and
137Cs
radioactivities with fuel
FIG. 21.7. Radioactivity of fission products per kg IHM in spent PWR fuel at 33 MWd/kg
burnup. Inflexion points indicate the existence of several radioisotopes.
595
15). As an example, 134Cs is not formed to any appreciable extent in fission because it is
shielded by the stable 134Xe. Hence, no 134Cs is normally observed in the remains after a
nuclear explosion in the atmosphere. However, primary fission products in the A 133 isobar
chain have time to decay to stable 133Cs during reactor operation and 134Cs is produced by the
reaction 133Cs (n,() 134Cs. Given the cooling time, the ratio between the decay rates of 134Cs
and 137Cs can be used to estimate the burnup of fuel from a given reactor, see Figure 21.6.
Using effective cross-sections and yield values the amounts and radioactivities in Figure 21.7
and Table 21.2 were calculated. It is seen that Xe, Zr, Mo, Nd, Cs, and Ru, which are the
elements formed in largest amounts in thermal fission (both by mole percent and by weight),
constitute about 70% of the fission product weight after a cooling time of 10 y.
At cooling times 10!1000 y the activities of 90Sr and 137Cs (with daughters) dominate among
the fission products. Later the fission product activity is due to very longlived nuclides of low
activity (Figure 21.7, Table 21.3). 131I, which for very short cooling times is one of the most
hazardous FPs because of its affinity to the thyroid gland, has practically
Fig. 21.8. Radioactivities of actinides and radium per kg IHM in spent PWR fuel after 33 MWd/kg burnup (see Tab.
21.2). Inflection points indicate the presense of several radioisotopes of the element.
596
Table 21.3. Some long-lived radionuclides produced in fission; tcool 10 y, data as in Table 21.2.
Nuclide
Half-life
(years)†
Decay; $!
(energy MeV)‡
Thermal fission
yield(%)††
activity (TBq/t U)
___________________________________________________________________________________________________________
79
Se
Kr
87
Rb
90
Sr 6 90Y 28.5
93
Zr 6 93mNb
99
Tc
107
Pd
126
Sn 6 126m2Sb
* `
9
* 126m1Sb
*
9
.6 126Sb
129
I
134
Cs
135
Cs
137
Cs 6 137mBa
151
Sm
154
Eu
155
Eu
85
<6.5×104
0.1509
10.72
0.6874
4.8×1010
0.2823
0.5462+2.2815
5.772
1.5×106
0.0905
2.13×105
0.2936
6.5×106
0.033
~1×105
0.368+3.670
0.0443
1.318
2.558
1.57×107
2.062
3.0×106
30.0
90
8.8
4.96
0.757
0
6.536
6.183
0.4196
0
0.0320
0.192
2.0585
0.205
0.5134
0.0763
1.9689
0.2527
0.0151
183
7.7×10!7
2180
6.375
6.074
0.147
0.0536
0.068+0.028
0.484
0.0042
0.029
0.0012
201
0.0098
3060
12
169
59
___________________________________________________________________________________________________________
†
Only for the longer lived mother nuclide. ‡ Decay energy, not particle energy (see decay schemes).
Thermal fission of 235U (fission of 238U, fission of Pu isotopes and n,(-reactions are important effects in a
nuclear reactor).
††
disappeared after a cooling time of 6 months. The A = 95 isobar chain, which is formed in
highest yield (6.52%), leads to 95Zr (t2 64 d) 6 95Nb(t2 35 d) 6 95Mo (stable).
From Table 21.2 it is seen that only stable isotopes remain for some fission elements at tcool
$ 10 y (Ga, Ge, As, Br, In, Xe, La, Nd, Tb, Dy and Er), while some others are of very low
activity (Zn, Se, Rb, Mo, Pd, Ag, Sn, I, Gd, Ho, and Tm). Some of the more active ones after
10 y have disappeared almost completely at 100 y (T, Sb, Ce, Eu, Pm, Ru, Rh, and Kr),
leaving essentially only the 90Sr-90Y and 137Cs-137mBa pairs and 151Sm.
21.3.2. Actinides
Neutron capture and $-decay lead to the formation of higher actinides. This is illustrated in
Figs. 16.2, 16.3, 19.5 and 19.7. 239Pu and 241Pu also fission, contributing significantly to the
energy production (Fig. 19.8). Truly, all plutonium isotopes lead to fission, since the n-capture
products and daughters, 241Am, 242Am, 243Cm, 245Cm, and most other actinides are either
fertile, fissible or fissile.
Fast neutrons in the reactor induce (n,2n) reactions, e.g. 238U (n,2n) 237U ($-) 237Np (n,()
238
Np ($-) 238Pu (n,() 239Pu (see Fig. 19.5), as well as fast fission. 237Np is also formed through
the reaction 235U (n,() 236U (n,() 237U ($-) 237Np. These reactions are the main sources of
neptunium and of 238Pu.
597
TABLE 21.4. Production of plutonium (kg/MWey) in various reactor types
Reactor type
Total Pu
Fissile Pu
________________________________________________________________________________________
Light water reactors
Heavy water reactors
0.51
Gas-graphite reactors
0.58
Advanced gas-cooled reactors
Liquid metal fast breeder reactors1.35
0.26
0.18
0.25
0.43
0.22
0.13
0.7!1.0
Many actinides formed are fissile but have a short half-life. The ratio, x, of the amount
fissioned to the amount decayed or reacted with neutrons at constant flux and steady state is
given by
x = N Ff / [8 + N (Ff + Fn,()]
(21.6)
where N is the neutron flux, Ff the effective fission cross-section, Fn,( the effective crosssection for radiative capture and 8 the decay constant. As can be seen from (21.6) a higher
fraction of the nuclide will fission in a very high neutron flux (x 6 Ff/(Ff + Fn,() when N 6 4),
whereas most of it will disappear by decay in a low flux (x 6 0 when N 6 0). As an example,
practically all 238Np formed by 237Np(n,() 238Np will fission when NFf o 8 (Ff = 2070 b, Fn,(
~0 b and 8 = 3.8×10!6 s!1). Hence, the buildup of many higher actinides is less efficient in
a high flux reactor than in a reactor with a more moderate neutron flux although the total
buildup rate might still be higher in the higher flux. This is accentuated in fast breeder reactors
where the combination of a very high neutron flux and a hard neutron spectrum (increased
effective Ff for fissible nuclides) strongly reduces the amount of higher actinides formed (at a
given burnup) compared to a thermal reactor.
From the fission and capture cross-sections, and half-lives the radioactivity of each actinide
element in one kg spent fuel (and radium) has been calculated and shown in Fig. 21.8 for a
PWR UO2 fuel with a burnup of 33 MWd per kg IHM. Due to the use of a log-log scale in
Figure 21.8, the decay curve of any single nuclides is s smooth curve bending downwards.
Inflexion points indicate the existence of several radioactive isotopes of the same element with
different half-lives.
The total amount of plutonium formed in various reactors is given in Table 21.4. The old
gas-graphite reactors and heavy water reactors are the best thermal plutonium producers. They
have therefore been used in weapons fabrication. The fast breeder reactor is also an efficient
Pu producer. Whereas thermal reactors (except at very low burnup) produce a mixture of odd
and even A Pu isotopes, a fast breeder loaded with such a mixture, by a combination of fission
and n-capture increases the relative concentration of Pu isotopes with odd A in the core and
produces fairly pure 239Pu in the blanket. Hence the combined Pu product from core and
blanket elements has a much higher concentration of fissile Pu isotopes than plutonium from
a thermal reactor (the fast breeder not only produces more Pu than it consumes but also
improves Pu quality, i.e. increases the concentration of fissionable isotopes). The LWR and
AGR are the poorest plutonium producers.
Composition of spent fuel varies somewhat with cooling time, e.g. the amount of 241Am
increases because it is the daughter of 14.4 y 241Pu; 116, 280 and 579 g/t IHM of Am after
598
FIG. 21.9. Average energy per decay in spent PWR fuel (fuel data according to Table 21.2).
0, 3 and 10 years, respectively.
TABLE 21.5. Decay heat from unpartitioned fuel, fission products and actinides (Basic data as
in Table 21.2)
Decay heat (W/kg initial U)
________________________________________________________________
Cooling time Total
Fission products
All actinides
___________________________________________________________________________________________
1d
90 d
180 d
1y
5y
10 y
100 y
103 y
104 y
105 y
106 y
193
30
19
10.8
2.1
1.4
0.32
0.054
0.013
0.0010
0.00038
146
29
18
10.3
1.9
1.2
0.13
0.000021
0.000019
0.000012
0.0000009
47
1.1
0.81
0.46
0.18
0.19
0.19
0.054
0.013
0.0010
0.00038
599
21.3.3. Decay heat and physical properties
As the radioactivity of the FP and actinides decreases by time, so does the energy absorbed
in the shielding material (and by self-absorption) which is seen as decay heat. Table 21.5 gives
data on the decay heat with contributions separately for the fission products and the actinides.
Figure 21.9 shows the variation of the average energy per decay with time. The maxima and
minima in the total and actinide curves in Figure 21.9 are caused by the presence of both "- and
$,( emitters with different activities and half-lives and also by the evolution in some decay
chains. For cooling times > 103 y the decay heat from the actinides and their daughters
dominates.
The decay heat is considerable at short cooling times due to the very high decay rate (see Fig.
19.15). Before unloading spent fuel from a reactor, the used fuel elements are first allowed to
cool in the reactor by forced circulation. Within a few weeks they are then transferred under
water to the cooling basin at the reactor site for an additional cooling time, usually 6!12
months, after which they may be transferred to a central spent fuel storage facility. In the
absence of such facilities, spent fuel elements can be stored in the reactor pools for many years.
During this time the radiation level and heat production decrease considerably.
21.4. Management of spent fuel
21.4.1. Transport of spent reactor fuel
The storage capacity of reactor pools is normally several years' production but can be
increased by adding neutron absorbers to the storage racks. Eventually the fuel assemblies must
be transferred in special transport flasks to (interim) storage sites, sent for reprocessing or to
final disposal.
The loading of used fuel the assemblies in the transport flask requires shielding and remote
handling, and the heat continual cooling. Therefore, almost all operations are carried out under
water. Because each transport is expensive the transport flasks are designed to carry several
assemblies of different types. A 30 t (gross weight) flask may carry 4 PWR or 9 BWR
assemblies (~1 t U), a 100 t flask ~12 PWR or ~30 BWR assemblies (~6 t U). The inner
cavity of such flasks is normally surrounded by a neutron absorbing shield; see Figure 21.10.
They have shock absorbers, and sometimes cooling fins on the outside. A filled 100 t flask
with 1 year old fuel develops ~60 kW heat; the design cooling capacity of the flask is ~100
kW. The cavity in some flasks is filled with water because the flasks are loaded and unloaded
under water, and the water functions as heat conducting and neutron absorbing material (several
actinides decay by spontaneous fission; ",n-reactions occur with light target atoms). Figure
21.10 shows a fuel cask for dry transport of 7 PWR or 17 BWR assemblies by special trucks,
by rail, or by sea in special RO-RO type ships.
The flasks are designed for exceedingly severe treatment: free fall from 9 m onto a concrete
floor, 30 min gasoline fire (~800EC), submersion into 15 m of water, etc., without being
damaged.
600
FIG. 21.10. Transport cask TN-17 for 7 PWR or 17 BWR fuel elements; N 1.96 m, length 6.15
m, empty weight 76 tons (SKB, Sweden).
After loading a flask the outside of the flask must be decontaminated; often it is quite difficult
to remove all activity, and a removable plastic covering is used. Before unloading, water filled
flasks are flushed to remove any activity leaked from the fuel or suspended crud, which is
present on the used fuel elements from BWRs and most PWRs. Further decontamination is
carried out after the flask has been emptied.
Dry air filled flasks are used in many countries. They are either cooled by filling with water
before unloading (the escaping steam is collected and condensed) or unloaded hot and dry by
remote handling. In the former case, further handling is the same as for water filled flasks.
Shielded transport is also required for solidified high level waste, hulls, plutonium containing
material, and for some intermediate and "-active waste. Special containers are used for each
type. As an example, in the United Kingdom plutonium containers made of wood and cadmium
are limited to carrying 10 kg Pu; the container weighs 175 kg, is 1.3 m high, and 0.8 m in
diameter.
21.4.2. Interim storage facilities
In the once-through fuel cycle, or with a limited capacity for reprocessing, separate spent fuel
element interim storage facilities become compulsory. These facilities are either of the wet pool
type (Czech Republic, Finland, Germany, Japan, Sweden, United Kingdom) or
601
of the dry vault type (France, Germany, United Kingdom). The pool types consist of large
water filled basins containing a geometrically safe (with regard to criticality) lattice of stainless
steel racks for the spent fuel elements. The water is circulated for cooling. In general such
pools are either located in a building at ground level or underground at 30!50 m depth.
Zircaloy clad oxide fuel elements can be stored for decades in storage pools with very little
risk of leakage. Metal fuels, especially those canned in magnesium or aluminum alloys, are less
resistant and should not be stored as such in this manner for a prolonged time. The corrosion
resistance of aluminum or magnesium clad fuel can be improved by electrolytic treatment
yielding a protective oxide layer.
Some of the stringent requirements on the storage pools are: keff < 0.95 (even if unused fuel
elements are introduced), earthquake safety, no possible water loss, water level automatically
kept constant, adequate leakage and radiation monitoring systems, water temperature < 65EC,
acceptably low radiation level in working areas, etc.
Dry vault storage may also be used. The fuel elements are stacked horizontally or vertically
in concrete pipes which allows cooling of the fuel elements by air convection (forced or
natural).
For long term storage (o10 y) the fuel elements must be recanned. For this purpose single fuel
elements or bundles are placed in cylindrical containers, and the void is filled with some
suitable material like lead, which has good heat conductivity and also provides some radiation
protection. Depending on the external condition at the final storing place (humidity,
temperature, etc.) the canisters are surrounded by an additional container to improve the
lifetime of the fuel elements, which ! preferably ! should not be exposed to the biosphere until
all radioactivity has disappeared.
21.5. Alternative fuel cycles
Fission energy can be obtained from uranium, using the uranium once-through option and the
uranium-plutonium fuel cycle, and from thorium, by the thorium-uranium fuel cycle. Each fuel
cycle offers a number of alternative routes with respect to reactor type, reprocessing, and waste
handling. Although the uranium based cycles are described with special reference to light water
reactors, the cycles also apply to the old uranium fueled gas cooled reactors.
21.5.1. The uranium once-through (UOT) option
The heavy arrows in Fig. 21.1 indicate the steps in the nuclear fuel cycle presently used on
a large commercial scale. The cycle stops at the spent fuel interim storage facility; from here
two alternative routes are available, one leading to the uranium-plutonium fuel cycle
(reprocessing of the spent fuel elements, as described in the next section) and another leading
to final storage of the unreprocessed spent fuel elements. The latter is referred to as the
once-through fuel cycle (UOT) option.
In the UOT option, the energy content of unused 235U, fertile 238U, and fissile and fertile
plutonium is not retrieved for future use, and the "waste" contains large amounts of these
602
and other "-emitting nuclides. At present this is the cheapest option and also withholds
plutonium from possible diversion to weapons use.
In 1994, only three countries (Canada, Sweden, USA) have decided to use the UOT strategy,
while all other nuclear power countries either were undecided or had elected to use the
reprocessing strategy. The technical aspects of the UOT option are further discussed in §21.13.
21.5.2. The uranium-plutonium (U-Pu) fuel cycle
"Cycle" infers to some mode of recirculation of material. The term "fuel cycle" was originally
used for the steps in which fissile and fertile material was isolated from used fuel elements
(reprocessing) and returned to the front end of the process for use in new fuel elements, see
Figure 21.1. The proper time to start reprocessing is a balance between loss of fissile material
(241Pu t2 14.4 y), economic interest loss on unused fissile and fertile material and storage costs
of the unprocessed fuels, on one hand, and, on the other hand, savings due to simplified
reprocessing and waste handling. Originally a 180 d cooling time was considered appropriate
(this time has been said to have been used in military programs). Presently the average cooling
time for commercial fuel elements is 7!10 y because of lack of reprocessing capacity.
The fissile fractions in spent LWR fuel elements amounts to ~0.9% 235U and 0.5!0.7%
239+241
Pu. By recovering these and returning them to the LWR fuel cycle the demand for new
uranium and enrichment services is reduced by ~30%. The uranium recovered may either be
re-enriched and used in normal uranium oxide fuel or blended with the plutonium recovered to
form mixed oxide (MOX) fuel elements (§21.1). MOX fuel can also be made from recovered
plutonium and depleted uranium. MOX fuel elements for LWRs contain up to 5% 239Pu+241Pu.
Many tons of plutonium have already been used as MOX fuel in LWRs.
The re-enrichment of recovered uranium leads to a small contamination of the enrichment
plant by 232U, 233U and 236U, which are introduced as part of recovered uranium. Today, most
batches of enriched uranium contain small amounts of these uranium isotopes.
In this cycle "old" plutonium from earlier production is successively exposed to increasing
neutron fluency, which changes its isotopic composition. This is shown in Table 21.6 for both
virgin plutonium and virgin + recycled plutonium. The fissile fraction (239Pu+241Pu)
decreases, requiring successively increased plutonium fractions in the MOX elements. The
concentration of the most toxic isotope, 238Pu, increases substantially. Its short half-life (88 y)
results in a high specific radioactivity which increases the heat evolution and radiolysis of the
reprocessing solutions. 238Pu, 240Pu, and 242Pu all decay partly through spontaneous fission,
with the emission of neutrons (~2 × 106 n kg!1 s!1). This makes the MOX elements more
difficult to handle. (All these changes also make the plutonium less suitable for weapons use.)
241
Pu has a critical mass of about half that of 239Pu, so as 241Pu builds up, criticality risks
increase. The buildup of other actinides by recycling high mass isotopes in LWR fuel further
increases reprocessing, handling, fuel manufacturing and waste problems.
Another possibility is to recycle only uranium, while the plutonium is left with the waste. This
would produce a waste with a very high concentration of plutonium, except for highly
603
TABLE 21.6. The isotopic composition of plutonium (weight %) as function of number of recycles
as MOX fuel in a PWR with a burnup of 33 000 MWd/t IHM per cycle.
236
238
239
240
241
242
244
Pu
Pu
Pu
Pu
Pu
Pu
Pu
Fraction left
__________________________________________________________________________________________________________________
Recycles
0
1
2
3
4
5
6
7
8
9
7.2×10!7
4.6×10!8
6.6×10!8
9.2×10!8
1.1×10!7
1.2×10!7
1.2×10!7
1.2×10!7
1.2×10!7
1.2×10!7
1.5
4.7
5.9
4.4
2.9
2.3
2.1
2.1
2.1
2.0
56.6
32.2
22.7
19.8
18.2
17.6
17.4
17.4
17.4
17.3
26.0
33.7
25.5
20.2
19.9
20.3
20.4
20.5
20.4
20.4
10.8
10.0
7.9
5.0
3.9
3.6
3.6
3.5
3.5
3.5
5.2
19.5
38.1
50.6
55.1
56.1
56.3
56.3
56.1
55.8
0.0004
0.0013
0.010
0.020
0.040
0.080
0.16
0.29
0.53
0.96
1.00
0.47
0.23
0.12
0.066
0.036
0.020
0.011
0.006
0.003
__________________________________________________________________________________________________________________
The recycling is assumed to be in the form of 5% Pu and 95% depleted U.
enriched uranium fuel; the weight ratio of plutonium to fission products would be 10:35 (kg/t
U, Table 21.2).
The reuse of plutonium in LWRs would only be temporary if fast breeder reactors would
become common (Fig. 21.1). FBRs are designed with a core, containing ~15% fissile Pu and
~85% 238U (as depleted uranium) in the form of mixed oxides or carbides surrounded by a
blanket of depleted uranium. In such a fast reactor, the presence of 235U is undesirable as it
reduces the neutron energy considerably by inelastic scattering. The actinide and fission product
contents in discharged FBR fuel (core and blanket) and discharged LWR fuel are roughly the
same on a GWey basis. Also the masses of total discharged fuel per GWey are comparable
because the high burnup in the core is balanced by a very low burnup in the blanket. Breeding
occurs only in the blanket (cf. §20.3). The burnup in the core elements is ~3 times higher than
in LWRs, and the fraction of fission products is also 3 times larger. Since only a small part of
the plutonium is burnt and the remainder has a high content of the fissile Pu isotopes, the used
core fuel elements retain a high economic value, making it desirable to reprocess them after a
short cooling time. Used FBR core elements may have a tenfold greater specific radioactivity
than spent LWR fuel elements at the time of reprocessing and a much higher content of
plutonium. Hence, criticality risks require a special reprocessing plant for core elements. The
FBR blanket elements are simpler to handle because of a lower content of fission products and
plutonium and they may be reprocessed in plants for LWR fuel elements. However, if the core
elements are sufficiently diluted with blanket elements such a mix may be reprocessed in a
"conventional" plant for LWR fuels.
In the FBR, 238U is consumed both by fission (i.e. energy production) and by 239Pu formation.
Because the "-value of 239Pu is 0.42, at least 70% (100(1 + ")!1) of all 238U is useful for
energy production in the U-Pu cycle. This value should be compared with the fairly small
fraction, # 0.7%, of the natural uranium which is used in the UOT cycle (taking enrichment
also into account), or # 1% in the LWR MOX fuel recycle. The FBRs not only increase the
useful energy of natural uranium by a factor of ~100, they also make
604
it possible to burn current stockpiles of depleted uranium amd also make it economic to mine
low grade uranium ore, vastly extending the available uranium resources.
In 1994, 15 nuclear energy countries had decided to use reprocessing as part of their strategy
for spent fuel management (Argentina, Belgium, Brazil, Bulgaria, China, Finland, France,
Germany, Hungary, India, Japan, the Netherlands, Switzerland, Russia and the United
Kingdom).
21.5.3. The thorium-uranium (Th-U) fuel cycle
Nuclear energy cannot be produced by a self-sustained chain reaction in thorium alone because
natural thorium contains no fissile isotopes. Hence the thorium-uranium cycle must be started
by using enriched uranium, by irradiation of thorium in a uranium- or plutonium-fueled reactor
or by using a strong external neutron source, e.g. an accelerator driven spallation source.
Fertile 232Th can be transformed into fissile 233U in any thermal reactor. The reactions in
232
Th irradiated by neutrons are given in Fig. 20.3. Of the thermally fissile atoms 233U has the
highest Ff /Fn,( ratio, i.e. highest fission efficiency. The 0-value is high enough to permit
breeding in the thermal region. Capture of neutrons in 233U is not a serious drawback as a
second capture (in 234U) yields fissile 235U, but reduces the breeding gain because two neutrons
are consumed without a net increase in fissile material. Since Ftot increases with neutron energy
(from ~7 b at 0.025 eV to ~26 b for reactor conditions; cf. also Fig. 19.3) the slightly harder
neutron spectrum in PHWR and GCR make such reactors the prime candidates for a
thorium-uranium fuel cycle together with the molten salt reactor (§20.3). The fuel may be
arranged in a core (233U) and blanket (232Th) fashion, or mixed fissile and fertile material as,
for example, in the HTGR prototype (High Temperature Gas-cooled Reactor) graphite matrix
fuels, or as a metal fluoride melt. The initial 233U must be produced from thorium in reactors
fueled with 235U (or 239Pu) or in special accelerator driven devices. After sufficient amounts
of 233 U have been produced, the Th-U fuel cycle may become self sustaining, i.e. thermal
breeding is established.
The advantage of the Th-U fuel cycle is that it increases nuclear energy resources considerably
because thorium is about three times more abundant on earth than uranium and almost as widely
distributed. In combination with the uranium fuel cycle it could more than double the lifetime
of the uranium resources by running the reactors at a high conversion rate (~1.0) and recycling
the fuel. Very rich thorium minerals are more common than rich uranium minerals. The
presence of extensive thorium ores has motivated some countries (e.g. India) to develop the ThU fuel cycle.
No full-scale Th-U fuel cycle has yet been demonstrated and reprocessing has only been
demonstrated on an experimental scale. The fuel cycle has to overcome the high activity
problems due to the presence of 228Th formed in the thorium fraction and 232U formed in the
uranium fraction (Fig. 20.3). The Th-U fuel cycle has a rather specific advantage over the UPu cycle in that its high active waste from reprocessing contains a much smaller amount of
longlived heavy actinides, and thus constitutes a smaller long term hazard. With regard to
nuclear weapons proliferation 233U is almost as good a weapons material as 239Pu and easier
to produce as a single isotope by continuous withdrawal of protactinium, since it is the decay
product of 233Pa (t2 27 d), see also §21.7.
605
21.6. Reprocessing of uranium and mixed oxide fuels
The main purpose of commercial reprocessing is:
(1) to increase the available energy from fissile and fertile atoms;
(2) to reduce hazards and costs for handling the high level wastes.
Two other reasons are sometimes mentioned:
(3) to reduce the cost of the thermal reactor fuel cycle;
(4) to extract valuable byproducts from the high active waste.
The 30% savings in natural uranium for LWR and similar reactors and the hundredfold energy
resource expansion for FBRs when reprocessing spent fuel, has already been discussed in the
previous section. The economic advantage of reprocessing depends on the cost and availability
of natural ("yellow cake") uranium, on enrichment and other front end activities, and on the
prevailing energy price (mainly based on fossil fuels). At present, cheap uranium is abundant.
The reprocessing plants at La Hague, France, have a total capacity of ~1600 tons IHM/y,
the Magnox reprocessing plant and the THORP (THermal Oxide Reprocessing Plant) at
Sellafield, UK, have capacities of ~1500 and ~850 tons IHM/y, respectively. A commercial
reprocessing plant under construction in Japan is designed to have a capacity of 800 tons
IHM/y. In Siberia, a large russian reprocessing plant is under construction. A number of
smaller, older plants are also in operation in several countries.
Figure 21.11 is a schematic representation of reprocessing of spent LWR fuel. The main steps
are: (i) the head end section, in which the fuel is prepared for chemical separation; (ii) the main
fractionation (partitioning) of U, Pu, and FP; (iii) purification of uranium ; (iv) purification of
plutonium; (v) waste treatment; and (vi) recovery of chemicals. These steps are described in
the following sections.
21.6.1. Head end plant
Figure 21.12 is a simplified drawing of one of the French oxide fuel head end plants at La
Hague. The flasks with the used fuel assemblies are lifted by a crane into water-filled pits,
where the flasks are unloaded and decontaminated. The assemblies are stored for a desired time
and then transferred to a shielded dismantling and chopping section. Some BWR fuels have end
parts, which can be mechanically dismantled; this is not the case for PWR fuels, in which the
end parts have to be cut off. The fuel pins are cut into pieces 3!5 cm long, either under water
or in air. At THORP, UK, and in the newest part of the La Hague plant, dry charging and
chopping are used.
The chopping is usually achieved with a shearing knife (cutter), but other techniques for
removing or opening up the zircaloy (or stainless steel) cans have been tried. Previously,
chemical decanning was used at some plants, e.g. Hanford and Eurochemic, but such
techniques increase the amounts of active waste considerably.
The chopped pieces are transferred to the dissolver unit, where the oxide fuel is leached by
boiling in 6!11 M HNO3 (the cladding hulls do not dissolve) in thick stainless steel
606
607
FIG. 21.12. Head end oxide fuel building at La Hague, France.
vessels provided with recirculation tubes and condenser. The hulls are measured for residual
uranium or plutonium, and, if sufficiently clean of fissile material, are discharged to the waste
treatment section of the plant. To improve the dissolution, some fluoride (# 0.05 M AlF3) may
be added to the HNO3. The F! forms strong complexes with some metal ions such as
zirconium, while its corrosion of the stainless steel equipment has been found to be negligible.
Soluble poisons, such as cadmium or gadolinium nitrate, are often added to the nitric acid to
assure the criticality control of the dissolution operation.
High burnup of fissile material leads to a high fission product content in the fuel elements
resulting in the formation of seminoble metal fission product alloys, which are insoluble in
boiling nitric acid. The insoluble material consists of mm sized metal particles of Ru, Rh, Tc,
Mo, and Pd. These metal particles usually contain negligible amounts of uranium and plutonium
and can be filtered as high level solid waste, HLSW.
When the fuel pins are cut, the volatile fission products contained in the gas space between
the fuel oxide pellets and the canning is released (mainly 85Kr, 3H, 129I, and 131I). These gases
are ducted to the dissolver off-gas treatment system. Gases released during dissolution are Kr,
Xe, I2, T2, THO, RuO4, CO2, minor amounts of fission product aerosols, and large amounts
of H2O, HNO3, and nitrogen oxides. Oxygen or air is fed into
608
TABLE 21.7. Specifications for reprocessed uranium and plutonium (From IAEA 1977)
Uranyl nitrate:
Uranium concentration 1!2 M
Free HNO3 $ 1 M
Impurities:
Fe, Cr, Ni # 500 ppm
Boron equivalents # 8 ppm†
Fission products # 19 MBq/kg U
"-activity (excluding uranium) # 250 kBq/kg U
Plutonium nitrate:
Plutonium concentration ~ 1 M
Free HNO3 2!10 M
Impurities:
Metallic # 5000 ppm
Uranium # 5000 ppm
Boron equivalents # 10 ppm
Sulfate # 1000 ppm
Fission products (t2 > 30 d) # 1.5 GBq/kg Pu‡
241
Am content (9 months after delivery to MOX-plant) # 5000 ppm
______________________________________________________________________________________________________
†
The equivalent values are B 1.0, Cd 0.4, Gd 4.4, Fe 0.0007, etc. The amount measured for each
of these elements multiplied by the factor indicated must not be more than 8 ppm.
‡ 95
Zr!Nb # 185 MBq/kg Pu.
the off-gas stream to allow recovery of part of the nitrogen oxides. The overall dissolution
stoichiometry is
UO2 + 2HNO3 + 2O2 6 UO2(NO3)2 + H2O
The gas streams pass to a condenser which reclaims and returns some nitric acid to the
dissolver. The noncondensibles are discharged to the off-gas treatment system.
When the dissolution is completed, the product solution is cooled and transferred to the input
measurement!clarification (filter and/or centrifuge) feed adjustment unit. At this point the
uranium is in the hexavalent state, and plutonium in the tetravalent.
21.6.2. Separation methods
The specifications for purified uranium and plutonium to be recycled are summarized in Table
21.7. Comparing these data with those presented before shows that at tcool ~1 y the fission
product activity must be reduced by a factor of ~107 and the uranium content in plutonium by
a factor of ~2 × 104. The large number of chemical elements involved (FPs, actinides and
corrosion products) make the separation a difficult task. Additional complications arise from
radiation decomposition and criticality risks and from the necessity to conduct all processes
remotely in heavily shielded enclosures under extensive health protection measures. As a result
reprocessing is one of the most complicated chemical processes ever endeavored on an
industrial scale.
609
The problem encountered by the chemists of the Manhattan Project in the 1940's was the
selection of satisfactory separation techniques. Advantage was taken of the relative stability of
the oxidation state of uranium (+6) and most fission products, and the redox lability of
plutonium (+3, +4, and +6). In the earliest process, only plutonium was isolated by
precipitating plutonium in the reduced state as PuF3 or PuF4 together with all insoluble FP
fluorides. This was followed by a second stage dissolving the precipitate, oxidizing plutonium
to the +6 state, and a new fluoride precipitation, leaving relatively pure plutonium in the
supernatant. In a final step plutonium was again reduced and precipitated as fluoride. This
principle was used for the first isolation of hundreds of kilograms of plutonium at the Hanford
Engineering Works, USA, with phosphate precipitation of Pu(+3) and Pu(+4), but not of
Pu(+6) which does not form an insoluble phosphate. Since Bi3+ was used as a carrier for the
precipitate, it is referred to as the bismuth phosphate process.
This principle of oxidizing and reducing plutonium at various stages of the purification scheme
has been retained in all subsequent processes. No other element has the same set of redox and
chemical properties as plutonium, though some elements behave as Pu3+ (e.g. the lanthanides),
some like Pu4+ (e.g. zirconium) and some like PuO2+
2 (e.g. uranium). Numerous redox agents
4+
have been used, e.g. K2Cr2O7 (to PuO2+
),
NaNO
2
2 (to Pu ), hydrazine, ferrous sulfamate,
4+
3+
and U (to Pu ), cf. §16.3.
The precipitation technique is not suitable for large-scale, continuous remote operations in
which both uranium and plutonium have to be isolated in a very pure state from the fission
products. It was therefore replaced in the late 1940's by solvent extraction in which the fuels
were dissolved in nitric acid and contacted with an organic solvent which selectively extracted
the desired elements. The technique has been mentioned in §§9.2.6, 9.4.3 and 16.3.3 but is
described in more detail in Appendix A.
The first solvent to be adopted at Hanford was methylisobutylketone ("MIBK" or "Hexone").
This solvent forms adduct compounds with coordinatively unsaturated compounds like the
actinide nitrates, e.g. Pu(NO3)4S2, where S represents the adduct molecule
Pu4+ + 4NO3- + 2S(org) 6 Pu(NO3)4S2(org)
The corresponding adduct compounds for 3- and 6-valent actinides are An(NO3) 3S 3 and
AnO2(NO3)2S2. These chemically saturated neutral compounds are soluble to different extent
in organic solvents like kerosene, and ! in the case of hexone ! by hexone itself. The process
using hexone is referred to as the Redox process.
In the United Kingdom, $,$'-dibutoxydiethylether ("dibutyl carbitol" or "Butex") was selected
as organic solvent; it forms the same kind of adduct compounds as hexone. Though more
expensive, it was more stable, less flammable and gave better separations.
Many other similar solvent systems, as well as chelating agents, have been tested.
Thenoyltrifluoroacetone (HTTA) was found to form strong complexes with the actinides (e.g.
Pu(TTA)4, UO2(TTA)2), which show very high distribution ratios in favor of organic solvents.
Though useful in the laboratory they were not found suitable for large scale commercial nuclear
fuel reprocessing. One of the most useful recent extraction agents for actinide separation is di-2ethylhexylphosphoric acid (HDEHP) which has found several industrial uses outside the nuclear
energy industry, e.g. separation of rare earth elements.
610
A drawback of hexone and butex is the need to use salting-out agents (salts like Al(NO3)3
added to the aqueous phase) in order to obtain sufficiently high extraction factors. Such salt
additions increase the liquid waste volumes. Further, hexone was unstable at high nitric acid
concentrations. All this led to the search for a better extractant.
Presently, tributyl phosphate (TBP) is the extractant in all reprocessing plants. It acts as an
adduct former and is normally used as a 30% solution in kerosene. It forms the basis for the
Purex process (Plutonium Uranium Redox EXtraction). TBP is cheaper than Butex, more
stable, less flammable, and gives better separations.
Other extractants, especially tertiary amines, have been tested for some steps in reprocessing.
The amines form organic soluble complexes with negatively charged metal complexes (used in
uranium purification, §5.5.3). The use in the basic Purex cycle of a secondary extractive
reagent can improve the decontamination factor1.
As alternatives to the aqueous separation processes, "dry" techniques have also been studied,
but none has been used on an industrial scale. Examples are the following:
(a) Halide volatility. Many FP and the high valency actinides have appreciable vapor
pressures; this is particularly true for the fluorides. In fluoride volatilization the fuel elements
are dissolved in a molten fluoride salt eutectic (NaF + LiF + ZrF4, 450EC) in the presence
of HF. The salt melt is heated in F2, leading to the formation of UF6, which is distilled; it may
be possible also to distill PuF6, though it is much less stable. The process has encountered
several technical difficulties.
(b) Molten salt extraction. The fuel is dissolved as above or in another salt melt. With a heat
resistant solvent of low volatility (e.g. 100% TBP), actinides and FP distribute themselves
between the two phases analogous to solvent extraction. This technique is of interest for
continuous reprocessing of molten salt reactor fuel or partitioning in an accelerator driven
transmuter (§20.3.1). An advantage is the higher radiation resistance of the nonaqueous system.
(c) Molten salt transport. The fuel is dissolved in a metallic melt, e.g. a molten Cu-Mg alloy,
which is in contact through a stirred molten chloride salt at about 800EC with another metallic
melt containing a reductant, e.g. a molten Zn-Mg alloy. Noble metal FPs are retained in the
Cu-containing melt, whereas U and/or Pu is collected in the Zn-containing melt. The less noble
FPs concentrate in the molten salt.
(d) Molten salt electrorefining. The spent fuel acts anode in a molten salt also containg a pure
metal cathode. By applying an electric field between the anode and cathode, material is
dissolved at the anode and deposited at the cathode. Careful control of the applied potential
makes it possible to obtain an extremely pure cathode product. The process was initially
developed for purification of plutonium metal alloys proposed as fuel for the Los Alamos
Molten Plutonium Reactor Experiment (LAMPRE).
(e) Molten metal purification. Metallic fuel elements can be molten and/or dissolved in molten
metals (e.g. a zinc alloy). In the presence of (deficient amounts of) oxygen, strongly
electropositive fission elements form oxides, which float to the surface of the melt as slag and
can thus be removed, while volatile FPs distill. The residual melt would mainly contain U, Pu,
Zr, Nb, Mo, and Ru ("fissium alloy") and can be reused in new fuel elements. This melt
refining technique has been tested on metallic breeder reactor fuel elements. Molten chlorides
have also been used to remove americium from molten plutonium metal scrap.
1
The decontamination factor is defined as (concentration before separation)/(concentration after separation).
611
21.6.3. Purex separation scheme
The distribution of uranium, plutonium, and some FPs between 30% TBP (in kerosene) and
aqueous solutions of varying HNO3 concentration is shown in Figure 16.8. DSr and DCs are n
0.01 for all HNO3 concentrations. The distribution of the trivalent lanthanides and actinides fall
within the Eu-Am area. Many fission products (most I, II, III, V, and VII-valent species) are
not extracted, i.e. D # 0.01; see §16.3.3 and Appendix A. Thus at high HNO3 concentration
Pu(IV), Pu(VI), and U(VI) are extracted but very little of the FPs. At low HNO3 concentration
the D-value for actinides of all valency states is n 1, and consequently the tetra- and hexavalent
actinides are stripped from the organic phase by dilute HNO3. This chemistry is the basis for
the Purex process.
The Purex process is presented schematically in Figure 21.11, where the solvent extraction
steps are within the dotted frame. Three purification cycles for both uranium and plutonium are
shown. High levels of beta and gamma activity is present only in the first cycle, in which
>99% of the fission products are separated. The principle of the first cycle is shown in Fig.
21.13. The two other cycles are based upon the same chemical reactions as in the first cycle;
the purpose is to obtain additional decontamination and overall purity of the uranium and
plutonium products. Each square in Figure 21.13 indicates a number of solvent extraction stages
of the particular equipment used: pulsed columns, mixer-settlers, etc. (see Appendix A).
In the first cycle > 99.8% U and Pu (in VI and IV state, respectively) are co-extracted from
3-4 M HNO3 into the kerosene-TBP phase, leaving >99% of the FP in the aqueous raffinate.
In the partitioning stage, plutonium is reduced to the III state by a solution containing a suitable
reductant, e.g. U(IV) nitrate or Fe(II) sulfamate; the plutonium is stripped to a new aqueous
phase and transferred to the plutonium purification section. Uranium, which as U(VI) (and
U(IV)) stays in the organic phase, is stripped by dilute HNO3 in a third stage. After
concentration by evaporation it is sent to the uranium purification section.
The uranium purification contains two extraction-stripping stages. Plutonium tracers are
removed by a reducing agent, e.g. U(IV):
3+
U4+ + 2Pu4+ + 2H2O 6 UO2+
+ 4H+
2 + 2Pu
U(IV) is preferred over Fe(II)-sulfamate in some plants as it avoids the introduction of foreign
substances. The final, concentrated uranium solution may be percolated through a column filled
with silica gel, which removes residual FP, particularly Zr-Nb.
The plutonium purification may be achieved by additional TBP extraction cycles. U(IV)
cannot be used as reductant in this part of the process. The final uranium and plutonium
products are nitrate solutions whose conversion to oxides, fluorides, etc., have been described
earlier (§5.5.3).
The chemical problems encountered in the solvent extraction are:
(i) The choice of diluent: A mixture of aliphatic hydrocarbons, and occasionally pure dodecane,
are most common; improper choice of diluent may lead to formation of a third liquid phase,
slow extraction kinetics, difficult phase disengagement (i.e. separation of the organic and
aqueous phases in the extraction equipment), etc.
612
FIG. 21.13. Flowsheet of the first purification cycle in the Purex process; TPH, tetrapropylene,
is a commercial dodecane. (From Musikas and Schulz.)
(ii) The choice of reductant for the reaction Pu(IV) 6 Pu(III): The use of U(IV) as reductant
for Pu(IV) introduces new uranium in the streams; Fe(II) sulfamate (Fe(SO3NH2)2) adds
objectionable inorganic salts to the aqueous high level waste; hydroxylamine (NH2OH) is a
kinetically slow-reducing agent; electrolytic reduction of Pu(IV) to Pu(III) is lacking
experience.
(iii) Solvent degradation: Radiation decomposes TBP into lower phosphates and butyl alcohol.
The main products are dibutyl phosphate (DBP, (BuO)2POOH) and monobutyl phosphate
(MBP, BuOPO(OH)2) which form strong complexes with many of the fission products as well
as plutonium. As these radiolysis products are formed the decontamination efficiency decreases
and losses of fissile material to the aqueous waste streams increase. The solvent is treated to
remove the degradation products prior to recycle in the process, e.g. by washing the TBP
solution successively with Na2CO3, NaOH, and dilute acid solutions.
To reduce some of these difficulties and to generally improve the efficiency of the U and Pu
extraction (leaving a waste almost free of these elements) as well as the decontamination
factors, other extractants than TBP have been suggested, for example dialkylamides,
R-CON-R'2 where R may be C3H7 and R' is CH2CHC2H5C4H9, "DOBA". The advantages of
this type and other proposed alternatives to TBP are that they are completely incinerable (while
TBP leaves a phosphate waste), the radiolytic products are not deleterious for the process
performance, and that no reducing agent is necessary to partition uranium and plutonium.
Finally, the ash from combustion could be leached and its content of actinides recovered and
returned to the process - further reducing the content of these elements in intermediate level
waste.
During the decontamination steps, acid streams containing small amounts of actinides and
fission products are produced. These streams are evaporated to concentrate the metal ions and
recycle them. Nitric acid is recovered from the condensates and recycled. Excess HNO3 may
be destroyed by formaldehyde. Fission product concentrates are routed to the aqueous raffinate
of the first extractor of the partitioning cycle which contains > 99% of the FP. This constitutes
the high level liquid waste (HLLW, or alternatively called HAW, high active waste). All other
liquid wastes can be subdivided into intermediate level waste
613
FIG. 21.14. Principle of a reprocessing plant with remote maintenance. 1, canyon with process
equipment; 2, control room; 3, feed preparations; 4, feed input; 5, cranes; 6, piping; 7, analytical
section.
(ILW or MLW) or low level waste (LLW). It is an important goal in all reprocessing operations
to reduce the amount of intermediate and low level waste streams as far as possible and to route
most of what remains to the HLLW stream. Waste treatment is discussed in §21.8 T.
21.6.4. Engineering aspects and operation safety
All operations in a reprocessing plant have to cope with the necessity of preventing nuclear
criticality and of protecting operations personnel and the environment from exposure to or
contamination by radioactivity. Thus all equipment has dimensions which are safe against
criticality, as, for example, annular or pipe shaped tanks for liquid storage instead of
conventional tanks, or are provided with neutron absorbers. All equipment is made of stainless
steel and is installed in concrete cells with wall thicknesses up to 2 m at the head end,
partitioning, and waste treatment sections. All operations are carried out in airtight enclosures
at reduced pressure relative to working areas.
In the event of failure of equipment within a radioactive area, three courses of action may be
taken: (i) switching to duplicate equipment, (ii) replacing or repairing equipment by remote
methods; remote maintenance, or (iii) repairing by direct maintenance after decontamination.
Plants were originally constructed for either completely remote maintenance or completely
direct maintenance. Figure 21.14 illustrates the principle of a plant for remote maintenance. All
equipment is installed in a large canyon with piping in a parallel corridor. The equipment can
be replaced by an overhead crane operated from a shielded room;
614
FIG. 21.15. (a) General layout of BNFL Magnox reprocessing plant at Sellafield, UK. (b) Cell
top of primary separation plant showing stirrer motors for mixer-settlers below.
615
malfunctioning equipment could be transferred to a decontamination and repair unit or to an
equipment "grave". With this philosophy conventional type chemical plant equipment can be
used, though redesigned for the remote replacement. Of particular importance in such plants
are good joints for piping, electricity, etc., for remote connection.
The British Nuclear Fuels Plc. (BNFL) Magnox and THORP plants at Sellafield, UK, are
designed for no maintenance, which in practice might mean either direct maintenance or remote
maintenance; Fig. 21.15. All kinds of common reasons for failure in chemical plants have been
eliminated or minimized by using welded joints and no moving parts. Thus there are no leaking
fittings, frozen valves, or stuck pumps. All welding is carefully controlled by ultrasound or
radiography. Active liquids are transported by gas (usually steam) jets or lifts, or by vacuum.
Liquid levels and volumes are measured by differential pressure gages and by weighing.
Samples for analysis are remotely withdrawn and analyzed in shielded boxes or glove boxes
depending on their activity outside the enclosure.
Even such systems may fail and dual equipment is therefore sometimes installed in parallel
cells. With the process running on the spare equipment, the failing equipment must be repaired.
This requires efficient decontamination both on inside and outside of the equipment, provisions
for which must be incorporated in the original design. This design requires dividing the plant
in a large number of cells so that the repair workers are protected from the radiation of the
functioning plant.
Remote maintenance is more expensive, but may be safer for the personnel and desirable from
the standpoint of continuity of operations because equipment replacements can be carried out
quickly and interruption of operation is relatively brief. Modern robot technology simplifies
such remote maintenance.
21.7. Reprocessing of thorium fuels
The thorium-containing fuels of present interest are only those of the kind used in HTGR and
HWR; in the future fuel from MSR-like (or other) transmutation devices may become
important. In the HTGR fuel elements the fertile ThO2 and fissile 235UO2 (or 233UO2) particles
are coated differently and embedded in a graphite matrix.
In case of the HTGR, the spent graphite fuel elements are mechanically crushed and burned
to eliminate the graphite matrix and pyrolytic carbon coating from the fuel particles. Leaching
permits separation of the fissile and the fertile particles because the fissile particles have a
silicon carbide coating which remains intact during burning and leaching, while the all-pyrolytic
carbon coatings on the fertile particles are burned away, allowing the oxide ash to be dissolved
by a leach solution, consisting of HNO3 and F!. The solution is clarified and adjusted to proper
acidity for solvent extraction (paragraph (i) below). The undissolved residue resulting from
clarification is dried and classified for further treatment (paragraph (ii) below). The burner
off-gas streams are passed through several stages of filtration, scrubbing, and chemical reaction
to remove the entrained and volatile fission products, as well as 14C containing CO2, prior to
atmospheric discharge.
In case of oxide fuel, it has to be cut and dissolved in a way similar to that described for
uranium fuels.
(i) The acid Thorex solvent extraction process is used to purify and to separate the 233U and
the thorium. Three solvent extraction cycles are used. In the first, the uranium and
616
thorium are coextracted by 30% tributyl phosphate (TBP) from ~5 M HNO3 and then stripped
into an aqueous phase. In the second cycle, the uranium and thorium are separated by
controlling the extraction conditions using # 1 M HNO3. The uranium is extracted and
processed by an additional solvent extraction cycle for final purification, while the thorium
remains in the aqueous raffinate stream. Following concentration and assay, the uranium is
ready for fuel refabrication. The partially decontaminated thorium is concentrated and stored.
(ii) The separated silicon-carbide coated fissile particles from the head end process for HTGR
fuel are mechanically crushed to expose the fuel and are burned to remove carbon and oxidize
the fuel material: the ash is leached to separate the fuel and fission products from the coating
waste. The 235U is separated from the fission products by solvent extraction using a Purex
flowsheet. An organic solvent containing only 3-5% TBP is used in order to avoid criticality
problems. A reductant such as Fe(II) sulfamate is added to the feed in order to force the small
amount of plutonium present into the aqueous raffinate stream. The waste streams are treated
as in the Purex process.
The process is plagued by both chemical and nuclear difficulties. The decay chain 233Th 6
233
U forms 27 d half-life 233Pa. For a complete decay of all 233Pa to 233U, the spent fuel
elements must be cooled for about a year. A still considerable amount of longlived 231Pa is
present in the spent fuel (about 1/2000 of the amount of 233U); protactinium complicates the
reprocessing chemistry and constitutes an important waste hazard.
The 233U isolated contains some 232U formed through reactions indicated in Figure 20.3. Since
the half-life of 232U is rather short and its decay products even shorter lived, a considerable
(-activity will grow in with the 233U stock, complicating its handling. Since the first decay
product is 228Th, some 228Th forms in the fuel elements, making the 232Th contaminated by this
isotope. Thus neither the 233U nor the 232Th produced can be free from (-activity.
21.8. Wastes streams from reprocessing
Fuel reprocessing generates a large variety of wastes which can be classified in categories
according to activity (low, medium and high), physical state (gas, liquid or solid package), or
decay characteristics (shortlived, longlived), each treated separately. The amounts of different
categories are given in Table 21.8; these wastes are treated internally at the plant and not
released to the environment. Because reprocessing plants vary considerably the amounts of
wastes produced also differs, especially for the liquid medium and low level categories. In this
section we briefly discuss the various waste streams appearing at the plant and the treatment
methods.
21.8.1. Gaseous wastes
Gaseous wastes originate mainly from the chopping and dissolution operations. In current
practice the volatile radionuclides are discharged to the stack after scrubbing with sodium
hydroxide and filtration through a special zeolite or charcoal filter. The hydroxide scrubbing
removes the acidic nitrous oxides which pass through the recombination unit above the
617
TABLE 21.8. Typical annual amounts of waste arising in a reprocessing plant (per t IHM and
3 years cooling time)
Type
Form
HLLW
MLLW
LLLW
liquid
liquid
liquid
HLLW
Fission products
Hulls
Ion/exch., precip.
Iodine
ML and LL "
Technical waste
conc. liquid
glass
concrete
bitumen
concrete
solid
concrete
Volume
$,(-activity
Pu-content
(m3)
(TBq)
(kg)
______________________________________________________________________________________________________________
5
5
100
0.5
0.15
0.8
0.8
0.03
0.1
3.2
105
105
105
400
~100
<0.01
~0.5
~0.05
0.05
0.05
0.01
0.008
!
~0.01
dissolver. The silver impregnated zeolite or charcoal removes remaining traces of iodine.
The most hazardous volatile constituents are the iodine and ruthenium fission products.
Though more than 95% of the iodine is volatilized in the dissolver (as I2, HI and HIO mainly)
most of it is caught in the off-gas scrubber and most of what remains is removed by the filters.
With these techniques the retention of iodine in the plant is >99.5%.
Ruthenium forms volatile RuO4 in the dissolver. Almost all RuO4 is retained in the gas
purification system. As an additional feature some plants use a steel wool filter (after the nitric
acid recombination) to catch the RuO4.
Of the noble gases, radioactive xenon has completely decayed after 1 y cooling, but krypton
contains 85Kr with 10.7 y half-life. This isotope is produced in appreciable amounts, and though
commonly it has been released to the atmosphere, this is no longer acceptable. Many processes
have been devised for krypton removal. Krypton in dry, clean air is effectively trapped on a
charcoal filter at cryogenic temperature; however, because of explosion risk (due to reaction
between radiolytically formed ozone and carbon), the favored process is condensation by liquid
N2 (krypton boils at !153EC) followed by fractional distillation. This removes >99% of 85Kr.
The krypton can be stored in pressurized cylinders until 85Kr has decayed (>100 years).
The amount of tritium released in reprocessing is considerable. In the chopping section it is
released as T2, but as HTO in the dissolver, where >90% of all tritium formed is present.
While T2 can be caught (particularly if chopping is done in air), a recovery of the HTO formed
in the dissolver would be expensive. Where it cannot be released to the environment the tritium
can be trapped. In one process, called Voloxidation, the chopped fuel elements are treated with
oxygen at 450-700EC before dissolution. Tritiated water is generated which should be relatively
free of ordinary water and consequently occupy a much smaller volume than tritiated wastes
do in present plants. Voloxidation may collect ~99% of the tritium present in unprocessed fuel.
Even when all tritium is released to the environment, some precautions against its spreading in
the plant are usually necessary in order to reduce the dose to operators. By proper design of the
first extraction cycle most of the tritium can be confined to this part of the plant.
618
TABLE 21.9. Composition of HLLW waste from Purex reprocessing of 1 t IHM LWR fuel with
a burnup of 33 000 MWd/t IHM and a cooling time of 3 years.
Component
Weight (kg) in original
Approx. molarity in
waste volume (~5 m3)
0.5 m3 concentrate
__________________________________________________________________________________________________________________
H+
NO3!
Fission products:
Corrosion products:
Group I (Rb, Cs)
Group II (Sr, Ba)
Group III (Y, Ln)
Zr
Mo
Tc
Group VIII (Ru, Rh, Pd)
Te
Others
Total fission products28.1
Fe
Cr
Ni
Total corrosion products
Phosphate (from TBP)
Actinides:
U (~0.5%)
Np (~100%)
Pu (~0.2%)
Am (100%)
Cm (100%)
Total actinides
Neutron poison (e.g. Gd)
1.4
900
2.94
2.37
10.31
3.54
3.32
0.77
4.02
0.48
0.35
~1.0
~2.4
0.046
0.041
0.15
0.076
0.068
0.016
0.078
0.0075
0.004
0.487
1.1
0.2
0.1
1.4
0.9
4.8
~0.44
~0.018
~0.28
~0.017
5.5
12
0.04
0.008
0.003
0.051
0.02
0.040
0.0037
0.00015
0.0023
0.00014
0.047
0.15
14
C is formed through the 14N(n,p)14C reaction in the nitrogen contained in the fuel elements.
It is released mainly as CO2 at the dissolution. Though only a small amount is formed (~40
GBq/GWe y for each ppm nitrogen in the fuel), its release to the environment, makes it the
dominating dose commitment of the fuel cycle back end (from reprocessing in 1989, 97 manSv
from 14C, and 98 manSv from all other nuclides). Presently 14CO2 is released to the
atmosphere, but techniques for its retention are available. It will probably finally be caught as
CaCO3, with a retention of ~80%.
21.8.2. Liquid wastes
The high level liquid Purex waste (HLLW) contains typically >99.5% of the FPs, <0.5%
of the U and <0.2% of the Pu present in the fuel as ~1 M HNO3 solution, see Table 21.9.
It is pumped to storage tanks as discussed further in §21.10. Total actual Pu losses during
reprocessing is presently # 0.2% of the feed.
The medium level liquid waste (MLLW) results essentially from evaporating various streams
from the chemical process, such as solvent clean-up, off-gas scrubbers, product concentration,
etc; Table 21.8. It may contain up to 0.5% of the uranium and up to 0.2%
619
of the plutonium processed. The radioactivity is usually <40 GBq/l (an average value is 4
GBq/l). The solutions also contain appreciable amounts of solids (e.g. NaNO3, iron, etc.). The
waste may be neutralized and is stored in steel tanks at the reprocessing site.
Liquid waste is generated in numerous places with activities <0.1 GBq/ m3. Such waste is
classified as low level. Some of these liquids may be clean enough to be released directly into
the environment. Others are cleaned by flocculation, ion exchange, sorption, and similar
processes. The general philosophy for liquid wastes is to concentrate all radioactivity to the next
higher level because the waste volumes decrease in the order LLLW > MLLW > HLLW.
Thus, in principle, the three kinds of wastes are reduced to two (HLLW and MLLW) and
cleaned aqueous effluent. The MLLW and residues from LLLW cleaning are treated as the
wastes of the nuclear power stations, i.e. concentrated and put into a disposal matrix such as
concrete or bitumen (see §20.4.3). At some coastal sites it has been the practice to release the
LLLW to the sea, with official permission. The nuclides of main concern are 3H, 90Sr, 137Cs,
106
Ru, and the actinides.
21.8.3. Organic wastes
The liquid organic waste consists of spent TBP diluent mixtures originating from the organic
solvent clean-up circuits and from the diluent (kerosene) washings of the aqueous streams (to
remove entrained solvent); in addition to degradation products of TBP and the diluent it contains
small amounts of actinides (mainly U and Pu) and FP (mainly Ru, Zr, and Nb).
This type of waste is disposed of by incineration or decomposed by hydrolysis and pyrolysis
leading to the formation of inactive hydrocarbons, which are distilled, and active phosphoric
acid, which is treated together with other aqueous wastes.
21.8.4. Solid wastes
High level solid waste (HLSW) originates at the dissolver. The hulls from the dissolution
contain activation products and small amounts of undissolved fuel (#0.1%). The dissolver
solution contains finely divided particles of undissolved seminoble metal alloys (Ru, Rh, Mo,
Pd, etc.). This suspension is treated by filtering or centrifugation prior to the solvent extraction.
Past practice in the USA and the USSR has been to put the HLSW in shielded containers which
are transported to and stored at a dry disposal site. In the future the same disposal is expected
to be used for the HLSW as for the solidified HLLW (§21.12).
Medium and low level solid wastes are produced at numerous places. They are divided in
various ways: combustible, noncombustible, "-bearing, non-"-bearing, etc., and treated
independently, when possible, to reduce volume. The wastes are then fixed in the disposal
matrix (§20.4.3). Table 21.8 gives the relative amounts of solid waste produced in
reprocessing. The final deposition of these wastes is further discussed in §21.13.
620
TABLE 21.10. Annual releases to the environment from some reprocessing plants
Source
Sellafield†
(TBq)
Half-life
(y)
Purex‡
(TBq/GWey)
__________________________________________________________________________________________________________________
3
H
14
C
85
Kr
Sr
95
Nb
95
Zr
99
Tc
106
Ru
129
I
90
131
I
137
Cs
144
Ce
241
Pu
U
airborne
liquid
airborne
liquid
airborne
airborne
liquid
airborne
liquid
airborne
liquid
12.33
5730
10.72
28.5
0.0958
0.175
2.13×105
1.02
1.57×107
222
1050
4.1
26000
600
150
150
180
810
0.074
0.022
30.0
0.781
14.4
4090
100
1800
11000 (kg)
41
643
2.0
0.54
12300
11
39
0.006
39
0.0007
0.032
0.002
13
__________________________________________________________________________________________________________________
†
Releases in 1978 only; 1.8 GWe produced from fuel reprocessed. ‡ Average for 3 plants normalized to 19851989; La Hague, France, Sellafield, UK, and Toka-Mura, Japan. (From UNSCEAR 1993.)
21.8.5. Environmental releases from reprocessing plants
As described in §21.8.1 relatively large amounts of radioactive gases with a low hazard index
(3H, 14C and 85Kr) are released to the environment, see Table 21.10. Other large non-gaseous
activities are contained within the plant, but small releases occur to the environment according
to limits set by the regulating authorities. These releases were rather large in earlier days, but
have been considerably reduced as improved procedures and equipment are coming into
operation, see Table 21.10. The authorized releases in the UK in 1978 were 104 TBq $-emitters
(excl. T) per year. The 137Cs release produced ~1000 Bq/l water in the middle of the Irish Sea
(~150 km SW Windscale), and could be traced as far away as in the Baltic Sea. The release
into the Atlantic in 1978 was #10 000 TBq. Not counting releases from plants in the US, USSR
and India, it is assumed to be <1000 TBq in 1992. It should be noted that the release of
radionuclides into the sea from nuclear power operations is dominated by the effluents from a
few reprocessing plants. It is, however, only a small fraction of the natural radioactivity in the
oceans.
21.9. Treatment and deposition of low and medium level wastes
It has become a common practice at reprocessing plants to store low level solid waste (and
sometimes higher level wastes) in trenches dug from the soil. These trenches, which
FIG. 21.16. (A) Storage of solid LLW (La Hague, France); (B) Storage of LLW and ILW
(Forsmark, Sweden); (C) Suggested repository for solid ILW and HLW (Belgium).
621
622
commonly are 5-8 m deep, are sometimes lined with concrete or simply have a gravel bottom.
For this purpose dry areas are selected (deserts, when available) or isolated areas with
controlled groundwater conditions with respect to water table depth, flow rate, and direction.
The disposed material normally should not exceed a dose rate of 1 mGy h!1 at 0.3 m distance,
or contain a specific activity exceeding 1 MBq kg!1. However, this varies, and may be a factor
of 10 higher or lower in some places. When the trench is full, it is backfilled with earth, after
which the surface dose rate usually is <0.01 mGy h!1. Trenches of this kind are used in the
USA, the UK, France, Russia, etc., where tens of thousands of cubic meters have been
disposed annually; Fig. 21.16(A).
Since some of the waste products are longlived, and the physical protection of the waste in
surface trenches is poor, radioactivity may ultimately leak into the groundwater, see Ch. 22.
Therefore, in many countries repositories for final deposition of solid intermediate and low
level wastes have been or are being built. The Swedish repository at Forsmark for ILW and
LLW from reactors, hospitals, research facilities, etc. has been in operation since 1988; Fig.
21.16(B). It is located in solid granite rock below 50 m of sea water at the Baltic Sea coast.
This underground site contains large concrete silos for the ILW and storage tunnels for ILW
and LLW containers. The space between the concrete walls and the rock is filled with clay
(bentonite). The repository is partly operated by remote control. Fig. 21.16(C) shows Belgian
plans for a mixed ILW and HLW repository located in a large clay deposit. Also in this
repository, concrete and clay are the main protective barriers. Geologic barriers are discussed
in §21.13.
21.10. Tank storage of high level liquid wastes
The main part of the HLLW is aqueous raffinate from the Purex cycle. It contains ~99.9%
of the nonvolatile FPs, <0.5% of the uranium, <0.2% of the plutonium, and some corrosion
products. For each ton of uranium reprocessed about 5 m3 of HLLW is produced. This is
usually concentrated to 0.5-1 m3 for interim tank storage; specific activity is in the range 107
GBq m -3 . The amounts of various elements in the waste and their concentration in 0.5 m3
solution is shown in Table 21.9. The HNO3 concentration may vary within a factor of 2
depending on the concentration procedure. The metal salt concentration is ~0.5 M; it is not
possible to keep the salt in solution except at high acidity. The amounts of corrosion products,
phosphate, and gadolinium (or other neutron poison added) also may vary considerably. Wastes
from the HTGR and FBR cycles are expected to be rather similar.
The HLLW stainless steel tanks have a volume of 50-500 m3. They are rather elaborate (Fig.
21.17); they contain arrangements for cooling (#65EC to reduce corrosion) and stirring,
removal of radiolytic gases, and for control of liquid level, pH, and radioactivity. The tanks are
usually double-walled, have heavy concrete shielding, and are often placed underground.
Storage in stainless steel tanks has been used in the last 40 y without failure. The philosophy
is that tank storage is only an interim procedure and will not last for more than a few years, but
the capacity for solidification has often been insufficient, so in practice tank storage has been
of much longer duration.
Tank storage has not been without failure. The mild steel tanks built in the 1940's at Hanford,
USA, were later (in the 1960's) found to leak due to corrosion. The waste
623
FIG. 21.17. Stainless steel tank for storage of HLLW.
nuclides seeped into the soil, some ultimately reaching ground water; NO3!, Ru and Cs were
found to have moved hundreds of meters, while Sr, Rare Earths and Pu were less mobile in
decreasing order.
In 1957 (possibly also in 1967) a serious accident occurred near Chelyabinsk (Kyshtym), south
of Ekaterinburg, in the former USSR, probably a chemical explosion between organic wastes
and nitric acid, in a high-level waste storage facility (tank or underground repository), leading
to the contamination of approx. 1600 km2 by 8×1016 Bq fission products. Local values as high
as 1010 Bq/m2 and values of 2×108 Bq/m2 for 90Sr and 137Cs have been reported. A large area
is now excluded to the public.
21.11. Options for final treatment of high level wastes
Many concepts are being studied to treat the high level liquid waste from reprocessing so that
the environment is protected against short and long term radiation damage. The final treatment
concepts are:
(a)dispersion to achieve environmentally acceptable concentrations;
(b) partitioning followed by
(i) nuclear transmutation;
(ii) disposal into space;
(iii) burial in a nonaccessible place;
624
© solidification followed by geologic deposition.
In principle, dispersion is only applicable for the gaseous and liquid wastes, which would need
negligible pretreatment. The limitations are practical (how efficient is the dispersion into air and
sea?), radiological (what radioactive concentrations are acceptable in air and sea?) and
political/legal (can it be permitted?).
The options (b) are not feasible for all high level waste products, and therefore has to be
limited to the most hazardous ones. It would require an isolation of these products, usually
referred to as partitioning, or fractionation, of the HLW.
Option (c) is the main route considered for high level wastes from reprocessing. Because of
its importance it is discussed in greater detail in a separate subsection (§21.12). Geologic
deposition is also the main choice for unreprocessed spent fuel elements.
21.11.1. Dispersion into sea and air
The main danger in release of radioactive waste is the risk of ingestion or inhalation. As the
hazard differs considerably between the various waste products, depending on their activity,
half-life, and biochemical properties, each radionuclide can be assigned a radiotoxicity value
(In), §18.13.6, defined by
and
Inw = A/ALI (man-years/kg spent fuel)
(21.7a)
Ina = A/DAC (m3 air/kg spent fuel)
(21.7b)
where A is the radioactivity (Bq) of a particular nuclide (cf. Fig. 21.7 and 21.8) and ALI and
DAC values are taken from Tables like 18.12; see also §18.13.6. Figure 21.18 shows the
radiotoxicity values for ingestion (Inw) and inhalation (Ina) according to the ALI and DAC
values recommended by the ICRP for the most hazardous fission products and actinides in 1
kg of unreprocessed spent PWR fuel at a burnup of 33 MWd/kg as a function of time. It is seen
that 90Sr and 137Cs dominate for the first 10 y, then followed by various actinides. The values
for spent BWR fuel are approximately the same.
The ocean water volume needed to contain all water soluble radionuclides and the volumes of
atmosphere needed to contain the gaseous waste products from all nuclear power in the world
at a level below the DAC and ALI values recommended by the ICRP for safe breathing and
drinking can be estimated. The basic data are the total toxicity values (Inw and Ina), which have
to be compared with the global (free ocean) water volume (1.4×1018 m3) and the atmospheric
volume (the troposphere volume up to 12 km is 6×1018 m3).
Taking all nuclides into account and multiplying by 15 000 000 (assumed to be the amount in
kg/year of spent fuel removed from nuclear power plants around the turn of the century) one
finds that the water volume needed is <0.1% of the ocean water volume. Thus in principle the
ocean capacity much exceeds the dispersion demand for quite a long time. Similarly, one finds
that the air volume needed is only a small fraction of the global air volume, assuming all
gaseous products to disperse evenly, i.e. the air volume can accommodate all nuclear power
gaseous waste products without approaching the DAC value; however, this is not the case if all
nuclides including the actinides were dispersed
FIG. 21.18. Radiotoxicity of the dominating nuclides in spent fuel; (A) for ingestion based on
ALI and (B) for inhalation based on DAC. Fuel data according to Table 21.2.
625
626
in the air. There is, however, no process for such a dispersal.
Considering that uniform dispersion is impossible, and that biological processes may enrich
some radionuclides, local concentrations would be expected leading to unacceptable doses by
exceeding the ALI and DAC values. Therefore only limited amounts are allowed to be disposed
of into air and the sea, and within strict rules on the kind and amount of nuclide and the
packaging prescribed by the London Convention (cf. §5.10.4).
The London Convention on the prevention of marine pollution by dumping of wastes (1972)
limits the amounts to 100 Ci y!1 for 226Ra, 10 Ci t!1 for other "-active waste of t2 > 50 y, 106
Ci t!1 for tritium, 100 Ci t!1 for 90Sr + 137Cs, and 1000 Ci t!1 for other $,(-waste at any site.
These figures are based on a dumping rate of not more than 100 000 t y!1 at each site. Dumping
is controlled by the IAEA. (N.B., 1 Ci is 37 GBq.)
In the diagram (Fig. 21.18(A)), the horizontal dashed lines indicate the Inw values for the
uranium involved in the fuel cycle. The three lines refer (a) to 7 kg natural U (i.e. the potential
hazard from the amount of ore that must processed in order to produce 1 kg U enriched to
~3.3% in 235U), (b) 1 kg enriched U, and (c) ~0.041 kg U (the amount 235+238U consumed
in 1 kg IHM at 33 MWd/kg); in the calculation of the In values all uranium daughters were
taken into account. It may be argued that when the radiotoxicity value for the waste goes below
such a line, the potential hazard is not greater than the natural material provided the external
conditions (Ex, eqn. (18.9)) can be made the same. The Inw line for HAW from reprocessing,
shown in Figure 21.19, and for 1 t natural U cross each other at a decay time about 105 y. Thus
if the HAW is evenly dispersed and equally well fixed as the original uranium and daughters
in the mine, it does not constitute any greater hazard than the uranium ore itself. Consequently,
it has been proposed to mix the HAW with cement and/or mine refuse, or, in a more elaborate
scheme, to mix it with rock-forming materials and convert it into a "synthetic rock"
(SYNROCK), and dispose of it in empty mines or natural underground cavities.
21.11.2. Partitioning
The term partitioning is used in two senses: in §21.6.3, to indicate steps in the Purex cycle
where U and Pu are separated; here, to indicate a separation of the most hazardous products
from the high level waste. Figure 21.19 shows that the hazard of the HAW is dominated by a
few fission products (mainly 90Sr and 137Cs) and the remaining actinides; the actinides in the
HLW consists of #0.5% U, #0.2% Pu and all Np, Am and Cm. In a successful partitioning
cycle, the U and Pu recovered will be recycled, leaving "the minor actinides" Np, Am and Cm,
which amount to about 0.8 g kg!1 y!1 spent fuel from a 1000 MWe LWR reactor; the
corresponding amount of Sr + Cs is about 3.7 g kg!1 y!1.
The amount of the minor actinides is small enough to make special disposal of them
interesting. If they could be removed from the HLW, the waste hazard would be considerably
reduced in time, see Fig. 21.19. Presently, research is directed towards the complete removal
of all longlived actinides from the reprocessing waste, followed by their "elimination" through
the alternatives described later.
These separation projects go under different names: in The US the approach is referred to
as the CURE (Clean Use of Reactor Energy), in which the TRUEX (TRansUranium EXtraction; also Truex) process is used; the Japanese have the OMEGA (Option for
627
FIG. 21.19. Relative radiotoxicity after 3 y cooling. From top: spent fuel, HLW, HLW after
removal of all Am and Cm, and after additional removal of all Sr and Cs.
Making Extra Gain from Actinides) program; the French the SPIN and ACTINEX (for
Separation Project Incineration Nucleair, and ACTINide EXtraction), etc. Many of these
projects are expected to run for decades before they can be realized.
It is a difficult task to isolate the higher actinides in the HLW, particularly to separate them
from the lanthanides, because these elements all are present in solution as trivalent ions of
similar size and therefore have very similar chemical properties. The separation methods utilize
their slightly different complex forming abilities in techniques such as solvent extraction, ion
exchange, and reversed phase partition chromatography. Three solvent extraction processes
have been run on a larger experimental scale:
(a) In the Reversed Talspeak process the extractant is di-2-ethylhexylphosphoric acid (HDEHP)
in a suitable aliphatic diluent. By adding lactic acid to the aqueous phase and adjusting pH to
2.5-3.0 the actinide and lanthanides are selectively co-extracted. The actinides can then be
stripped by an aqueous phase containing diethylenetriaminopentaacetic acid (DTPA) as complex
former and lactic acid as kinetic promotor, leaving the lanthanides in the organic phase.
(b) The Truex process is based on a carbamoyl organophosphorus extractant, abbreviated
CMPO (for octylphenyl-N,N-disiobutylcarbamoylmethylphosphine oxide, chemical formula
C8H7(C6H5)POCH2CON(CH2CH(CH3)2)). CMPO is used together with TBP in a mixture of
aliphatic hydrocarbons. The Truex flow sheet is shown in Fig 21.20. This process (like
628
the other ones) is based on a common principle: combining complexation of the species in the
aqueous phase with a selective extraction process or reagent.
A "complete partitioning" requires the removal of strontium and cesium from the HLW, as
90
Sr and 137Cs dominate the waste hazard during the first 400 years. If these are removed, the
waste becomes almost "harmless" (i.e. it will be quite easy to handle), see Figure 21.19.
Solvent extraction processes have been developed which effectively remove Sr and Cs. Such
process have to be integrated with the actinide removal processes for an efficient back end
cleaning procedure. Sr can also be removed from HLLW by the Srex process, in which it is
very selectively extracted by a macrocyclic ether (ditertiarybutyldicyclo-hexanone-18-crown-6
dissolved in n-octanol, also 99Tc is extracted). Recently another group of crown ethers, calixarenes, have shown a high selectivity for Cs: e.g. in 1 M HNO3, DCs is ~20, while DSr is only
0.01. Thus improved methods are being developed for the efficient removal of Sr and Cs from
high level waste.
The main point of partitioning the high level waste is that it shall lead to a safer waste (more
acceptable to the public) as well as a cheaper back end fuel cycle (to the advantage of the
nuclear energy industry). Koyama has analyzed the different waste handling options, Fig.
21.21, and concluded that full partitioning (actinides as well as Sr+Cs, Tc and I) will lead to
the cheapest fuel cycle.
21.11.3. Disposal into space
Assuming 20 t IHM spent fuel are removed from a 1000 MWe LWR annually, the partitioned
waste would amount to ~20 kg minor actinides and ~80 kg Sr+Cs annually. The payload of
modern rockets exceed several tons. Thus, even including the weight of packaging material,
it does not seem unreasonable to assume that the minor actinides from a considerable number
of LWRs (>10) could be transported by a rocket or other means to a location off the Earth.
Several different space trajectories have been considered. These include:
(i) a high Earth orbit (altitude 150 000 km); )v 4.15 km s!1;
(ii) transport to the sun; )v ~22 km s!1;
(iii) an inner solar orbit; )v ~15 km s!1;
(iv) solar system escape; )v ~52 km s!1.
)v is the incremental velocity required to leave a circular Earth orbit at 200 km and is a direct
indication of the size and propulsion energy of the rockets required. Vehicles that could be used
include existing rockets and the space shuttle.
A high Earth orbit has the advantage of low )v and possible later retrieval of the waste, but
it requires long term container integrity and very long orbit lifetime (not yet proven). Transport
to the sun or solar escape has the advantages that the waste is permanently eliminated from the
Earth, but a very high )v is required. The inner solar orbit has the next lowest )v but requires
a very stable orbit, so that it does not return to Earth.
Space disposal has an economic disadvantage since the cost of transportation is likely to be
very high, as the weight of shielding could create significant economic penalty. There
629
FIG. 21.20. Truex flow sheet for removal of minor actinides from HLLW (From Musikas and
Schulz).
are also requirements for capsule integrity to provide a reasonable degree of assurance of
survival in the case of an abort. At present, this is not considered to be a serious option.
However, many kilograms of 238Pu produced from 237Np have already been used as a power
source in deep space probes or unmanned landers (c.f. §6.9.3).
21.11.4. Remote deposition
It has been suggested that the most hazardous actinides be removed from the reprocessing
waste and stored separately. The advantage is (i) to eliminate all actinides from the HAW, so
that its hazardous potential follows that of 90Sr, 137Cs and 99Tc, by which the main hazard is
gone in about 400 years, and completely in about 100 000 y; (ii) to reduce the waste actinides
to a small volume, 1/100 to 1/1000 of that of the HAW (and, of course, even much less when
compared to the volume of the spent fuel elements), which will simplify the storage problem.
For example, such actinide waste could be stored uniquely in very deep bore holes in the
ground, eventually in the earth's molten interior. A similar procedure seems possible also for
the Sr+Cs fraction.
630
FIG. 21.21. Comparison of costs of various options for managing and disposing of Purex highlevel waste. (From Koyama.)
21.11.5. Transmutation
The fission products 90Sr and 137Cs can be transformed into shorter lived or stable products
by charged particle or neutron irradiation. Charged particle irradiation would be very expensive,
and irradiation by reactor neutrons would produce almost as much fission products as are
destroyed. Therefore the use of intense accelerator driven spallation neutron sources for
transmutation by n-irradiation has been suggested. If controlled thermonuclear reactors (CTR)
are developed, their excess neutrons could be used for 90Sr transformation, but less efficient
for 137Cs.
In the long term ($600 y) the actinides dominate the risk picture. Continuous neutron
irradiation of the actinides finally destroys all of them by fission (cf. Fig. 16.3). The annual
production of americium and curium is ~5 kg in a 1000 MWe LWR, but considerably less in
a FBR. Thus if pins of these elements are inserted in a FBR, more americium and curium is
destroyed than formed; it is estimated that 90% will have been transformed into fission products
after 5-10 y. In the future, CTRs could be used for the same purpose. As an alternative it has
been suggested to leave the americium and curium in the uranium returned in the LWR cycle.
Wastes from transmutation processes will contain some amount of longlived nuclides, thus a
safe final repository is still needed.
631
FIG. 21.22. Alternative flow sheets for HLLW solidification.
21.12. Solidification of high level liquid wastes
The solidification of HLLW, followed by geologic deposition, is presently considered as the
only realistic technique to create conditions for a safe long term disposal of HAW. The
objectives of solidification is to immobilize the radioactive elements and to reduce the volume
to be stored. The solidified product must be nondispersable (i.e. not finely divided as a
powder), insoluble, and chemically inert to the storage environment, be thermally stable, have
good heat conductance (this determines the maximum radioactivity and volume of the final
product), be stable against radiation (up to 1010 Gy), and have mechanical and structural
stability.
Figure 21.22 shows the options for solidification and encapsulation of HLLW. The first step
is usually a calcination in which nitrates are destroyed and all metals converted to oxides.
Thousands of cubic meters of HLLW have been solidified by fluidized bed calcination.
However, calcine has low leach resistance, low heat conductivity, and can be dispersed in air.
It is, therefore, only considered as an interim product.
Most countries have focused development work on the fixation of the active waste in
borosilicate or phosphate glass. Large continuous vitrification plants producing borosilicate
waste glass are located at La Hague in France and at Sellafield in the UK. Figure 21.23 shows
a rotating calciner which feeds a continuous borosilicate glass melter. Preformed and crushed
borosilicate glass is continuously added near the end of calcination process. In
632
TABLE 21.11. Characteristics of some solidified high level waste products
Property
Fluidized bed calcine
Phosphate glass
Borosilicate glass
__________________________________________________________________________________________________________________
Physical form
Granular, 0.3!0.7 mm
Bulk density (kg m!3)
1.0-1.7
Maximum weight % FP
50
Thermal conductivity (W m!1 K!1)
0.2-0.4
Leachability (20EC, kg m!2 d!1) 1-10
Maximum center temperature (EC)
550 ††
Maximum allowable
#
45
heat density (kW m!3) @ @
38
Monolithic
2.7-3.0
10!5-10!3 †
36
Monolithic
3.0-3.5
35
1.0
10!6-10!4 ‡
400 *
53
92
30
1.2
700 *
127
__________________________________________________________________________________________________________________
†
!2
!1
!2
!1
Devitrified (crystalline) glass has leach rates 10 -10 kg m d .
At 100EC the leach rate is ~10!2 kg m!2 d!1.
††
Because of risk for FP volatilization.
*
Devitrifies at higher temperature.
# Forced water cooling of a cylinder; N 0.3 m, surface temperature #100EC.
@ Natural air cooling of a cylinder; N 0.3 m.
‡
the melter the calcines and glass mixture is heated to 1000-1200EC, leading to the formation
of a homogeneous glass, which is poured into stainless steel cylinders. From a chemical
standpoint (stability) the glass can contain up to 30 wt% fission product oxides, but 20% is a
more normal value. Thus the waste nuclides in 1 t of LWR fuel can be contained in about 200
kg glass; the exact value depends on how long time the waste has cooled. For short cooling
times a high waste content in the glass could cause it to melt and possibly crystallize (at
~700EC), which may reduce corrosion resistance because of the larger internal surface
formed. The properties of some glasses are given in Table 21.11.
FIG. 21.23. Continuous HLLW calcination and borosilicate vitrification process.
633
The incorporation of the solidified waste into phosphate glass is an alternative to the use of
borosilicate glass. In this case the calcination step can be bypassed. The HLLW, together with
phosphoric acid, is evaporated and denitrated, and then fed to a continuous melter operating at
1000-1200EC, from where the molten glass flows into the storage pot. This process is more
corrosive and produces a glass somewhat inferior to the borosilicate. Further disadvantages are
that the glass recrystallizes at relatively low temperature (~400EC), and that the devitrified
(crystallized) phosphate glass exhibits a rather high leach rate, 0.01!0.1 kg m!2 d!1.
In the PAMELA process, earlier considered for use in Germany, granulated phosphate glass
is incorporated into a metal alloy matrix in a steel cylinder. This offers high chemical and
mechanical stability, as well as good heat conductivity (~10 W m!1 K!1 for a lead matrix; cf.
1.2 W m!1 K!1 for borosilicate glass). This decreases the central temperature and allows the
incorporation of larger amounts of radioactivity (up to 35% FPs) in a single cylinder. The high
heat conductivity makes it feasible to use short fuel cooling times (down to 0.5 y) and
diminishes the demand for interim fuel element storage basins or HLLW storage tanks. The
solidified waste glass is collected in stainless steel cylinders.
A typical waste canister may contain ~9% FPs, emit ~10 kW after 1 y, ~1 kW after 10
y, and ~0.5 kW after 40 y. They can be stored in air-cooled vaults or water-cooled pools.
Before final deposition the HAW containers are reconditioned, i.e. enclosed in an additional
canister of type described below. At the higher power level they are stored in water-filled pools.
After >10 y air-cooled vaults may be satisfactory; forced cooling is usually considered, but
with precaution for sufficient convection cooling in case of ventilation failure.
21.13. Deposition in geologic formations
A general consensus has developed that disposal of radioactive waste should be in the country
in which the nuclear energy is produced. Thus, even if the spent fuel elements are sent from
one state to another for reprocessing (like spent fuel from Japanese reactors being sent to the
Sellafield reprocessing plant in the United Kingdom), the producer state (Japan) must guarantee
its readiness to accept the processed waste. This condition has led to intensive national
investigations of the safest way to dispose of the high level waste, either in the form of
solidified HLLW, or spent fuel elements in the once-through option. In these studies geologic
repositories have been the prime consideration, e.g. crystalline silicate rock, clay layers, salt
domes, the sea bed, etc.
21.13.1. Properties of geologic formations
Many geologic formations are being studied for final storage of the waste canisters: rock salt,
crystalline rock (e.g. granite), volcanic tuff, clay, etc. The main requirements are:
(i) geologic stability, i.e. regions of very low seismic, volcanic, or other geologic activity;
(ii) absence of large fractures, holes, etc.;
(iii) impermeability to surface water;
634
(iv) negligible groundwater circulation with no flow-lines leading to nearby potential intake
sources;
(v) good heat conductivity;
(vi) of little interest to and as remote as possible from human activities.
There are large geologic formations, which are very old (many hundred million years) and
geologically stable, e.g. the Scandinavian and Canadian shields. These are usually fractured,
even if considerably large volumes can be found (106 m3) which are free of fractures. A nuclear
waste storage facility will, however, for reasons given below, cover rather large areas, several
km2 ; thus for crystalline rock fracture zones, which carry ground water, must be taken into
account. Rock salt and clay formations, on the contrary, are usually free of fractures: rock salts
are dry, but clay formations are usually wet, though the water migration through the formation
is extremely small (see below).
A limiting factor for geologic disposal is heat conductivity. If the thermal conductivity of the
geological deposit is too poor, the material in the waste canisters may melt and react with the
encapsulation, possibly destroying it. Because a safe disposal concept must rely on conduction
cooling, the relatively low heat conductivity of the geologic deposit makes it necessary to
disperse the waste material over a large volume. The thermal conductivity of ~3 W m!1 K!1
for granitic rock (compare rock salt ~9 and wet clay ~1) limits the waste heat density of the
rock wall to #20 W m!2. Therefore storage repositories would consist of long tunnels or deep
holes at large interspace. Three concepts are under investigation: (i) very deep holes (VDH,
up to 3000 m) in which waste canisters would be stacked, (ii) deep repositories (300-500 m)
with parallel horizontal tunnels of moderate length (#100 m), 20-100 m apart, upper part in
Figure 21.24, and (iii) deep repositories with a few very long tunnels (>1 km, VLH).
With a typical heat production of ~500 W per canister containing vitrified HAW (at 40 years
cooling) the wall heat density would be 6-7 W m!2. The internal temperature of a glass canister
would then never exceed 90EC, and the surface temperature (above the storage holes, Fig.
21.24) not more than 65EC. Slightly higher temperatures can be accepted for spent fuel
elements.
Such deep (geologically) repositories are safe against surface activities (including nuclear
explosions) except drilling or mining. In order to avoid such an occurrence after the repository
has been forgotten, it should be located in a formation of no interest to society; e.g., the
formation should not contain any valuable minerals. From this point of view, location in deep
seabeds has been suggested. Such a location can be chosen either deep in the bottom silt or in
a bedrock under the sea floor, or, possibly, in a tectonically active trench, where with time the
waste would be pulled down into the interior of the earth.
The risk for a land-based geologic repository is that a combination of more or less
unpredictable circumstances could lead to rupture of the protective barriers around the waste
matrix, followed by dissolution and transport of the most hazardous products by water to a
place where it can enter into the food chain. Therefore a clay layer (e.g. bentonite) between the
canister and the rock wall is suggested, partly to act as a mechanical buffer, so that even
considerable slippage caused by earthquakes would have little effect on the mechanical integrity
of the canister and partly to reduce water flow around the canister.
Most rocksalt deposits are very old. In many cases they rise in the form of a dome, the top
of which may be at a few hundred meters below surface level, and with a depth up to
FIG. 21.24. Principles of storage of HLW: upper part, Swedish concept; lower part, canister with
inserts for final storage of spent BWR fuel and spent PWR fuel (From SKB).
635
636
more than 1000 m. The dome is usually protected from groundwater by a calcite (CaCO3) cap.
Salt has been mined in such formations for centuries. The formation contains only
microcrystalline water and is extremely dry (otherwise it would have dissolved during the
geologic ages). Therefore, canisters emplaced in such formations are not believed to dissolve
(as long as no water enters through the hole mined through the calcite layer). The good heat
conductance allows emplacement of waste of higher power density than for deposits in clay or
granite. Since the salt is plastic, the holes and corridors, which are backfilled with crushed salt,
selfseal and no clay buffer is needed. The canisters would probably be crushed in the plastic
salt, and therefore could be irretrievable after a hundred years.
Granite and clay formations are percolated by groundwater except at depths which presently
are considered impractical. The water flow rate is given by Darcy's law
Fw = kp i S (m3 s!1)
(21.8)
where kp is the permeability (m s!1), i the hydrostatic gradient (m m!1), and S the flow cross
section (m2). A typical value for Scandinavian granite formations at 500 m depth is Fw =
2×10!4 m3 m!2 y!1, which gives a ground-water velocity of 0.1 m y!1, and a time of 5000 y
for the water to move 500 m. In this case the rock permeability is taken as 10!9 m s!1; rock
formations closer to the surface usually have higher permeabilities, but many formations with
much lower values are also known (~10!13).
Clays consist of small particles, usually <2 µm and with an average size of <0.1 µm, of
various minerals like quartz, feldspar, montmorillonite (a hydrated aluminum silicate with high
ion-exchange capacity), mica, etc. The overall chemical composition is mainly a mixture of
silica, alumina, and water. The small particles in the clay give rise to a very large surface area
(1 cm3 of particles of 0.1 µm diameter have a total surface area or about 60 m2) and
correspondingly high sorption capacity for (eventually) dissolved waste products. Although the
clay can take up large amounts of water (up to 70%) without losing its plasticity, the water
permeability is extremely low. For the sodium bentonite clay (~90% montmorillonite, plenty
available) considered in many projects, the permeability is 2×10!14 m s!1 (10% water in clay
compacted to a density of 2100 kg m!3). Such clay can be considered impermeable to
groundwater. Natural clays of this type occur in many countries (the Netherlands, Italy, etc.).
21.13.2. Waste conditioning before final storage
The basic philosophy in storing high levels of radioactivity in geologic formations is the use
of a multi barrier system to protect it against dissolution. These barriers are:
(i)
(ii)
The waste matrix itself, which is made highly insoluble; this is achieved by
vitrification of the reprocessing HLLW as described above, while the spent
unreprocessed fuel elements keep the radioactive products in a highly insoluble UO2
matrix.
The waste matrix (glass or UO2) is encapsulated in an "insoluble" canister.
(iii)
637
The canister is surrounded by a clay buffer in the case of a crystalline rock deposit.
Here we shall only discuss the encapsulation, for which numerous concepts have been
suggested.
(a) Vitrified waste. The stainless steel cylinders containing the vitrified radioactive waste are
not considered to provide sufficient protection against long term corrosion in the final
repository. Additional encapsulation in various corrosion resistant materials is therefore
suggested: lead, titanium, copper, gold, graphite, ceramics, etc. For example, in one proposal
canisters containing 170 l vitrified waste in a 3 mm thick stainless steel cylinder will be
surrounded by 100 mm lead in a casing of 6 mm titanium; the overall dimensions of the
cylinder will be 1.6 m long with a diameter 0.6 m. One such cylinder contains solidified HLLW
from reprocessing 1 t spent fuel from a 1000 MWe light water reactor; this corresponds to
about 20 cylinders per year.
(b) Spent fuel. Although UO2 dissolves more slowly than glass in groundwater, spent fuel
elements must be recanned before entering the final storage facility. For this purpose single fuel
elements or bundles are placed in cylindrical canisters. The lower part of Figure 21.24 shows
an example. The outer part of the canister may consist of iron, stainless steel, copper, etc.
selected to resist the repository environment (e.g. dry or wet). For storage in very long or deep
dry holes, rounded caps are likely to be used. Waste canisters for unchopped spent fuel
elements will be rather long, but it has also been suggested that the fuel is chopped as in the
German Pollux concept, Fig. 21.25, which is designed for multi-purpose use, both to fit all
kinds of spent fuel elements (from BWRs, PWRs, HTGRs etc) as well as vitrified high level
waste. It is a double-shell concept, consisting of leak-tight welded steel; between the two steel
walls is a corrosion resistent sheet of Hastelloy C4. The void is either gas filled (He) or filled
with some suitable material like lead, which has good heat conductivity and also provides some
radiation protection. Figure 21.24, shows the design of a spent fuel element canister where the
fuel elements are inserted in a cast iron cylinder which is surrounded by 50 mm of copper
metal. The thick metal layer has several purposes: (i) to simplify handling during deposition
operations; (ii) to protect its content from mechanical deformation; (iii) to reduce the surface
radiation to so low values that the radiolysis of groundwater will not contribute to canister
corrosion; (iv) to protect, as long as possible, the waste-containing matrix from corrosion. For
example, by using an outer copper wall, the lifetime of the canister is expected to greatly
exceed 1000 y for oxygen containing groundwater, and 104 y for reducing groundwater (the
lifetime of a 5 cm thick copper layer is now estimated to be >106 y). The dissolution of waste
canisters and leaching of waste is discussed in §22.10.
21.13.3. Repository projects
In one concept the canisters are stored in dry areas (e.g. a desert floor) after having been
surrounded by concrete. A space may be left between the inner cylinder and the surrounding
concrete wall to allow air convection cooling. As long as the climate stays dry the pillars will
erode very slowly and last for tens of thousands of years. This procedure also allows for easy
retrievability. An alternative encapsulation is achieved by using hot
638
FIG. 21.25. The proposed German POLLUX cask for final disposal of spent fuel and/or vitrified
high level waste in salt.
(~1500EC) isostatic (100-300 MPa) compression to surround the fuel bundles by a
homogeneous, dense ceramic material, like corundum (microcrystalline Al2O3) or graphite.
Since corundum and graphite are natural minerals, the long term resistance should be very high,
even against water.
The main concept is to store the vitrified HLW and the spent fuel elements irretrievably in
deep underground geologic formations, according to designs described above. Many such
formations are now being evaluated: rock salt (mainly Germany, the Netherlands and USA),
crystalline rock (Canada, Finland, France, Japan, Sweden, USA), volcanic tuff (USA), clay
(Belgium, Italy), etc. Cross sections of geologic repositories for HLW are shown in Figures
21.16(C) and 21.24. Also disposal into polar ice sheets (USA) and the seabed (UK, USA) have
been investigated, but are no longer considered feasible.
The most advanced projects are (i) at the abandoned Asse salt mine in Germany, where drums
with low and intermediate level activities have been deposited at 300 m depth in 1967-78; the
repository is now used for research; a final repository in salt at Gorleben is planned to be in
operation be the year 2010, (ii) the waste isolation pilot plant (WIPP) at
639
a bedded salt deposit in New Mexico at 600 m depth; the planned repository is currently
examined by the authorities for full scale operation; it will primarily be used for "-bearing
military waste; in Yucca Mountain extensive studies are being made for a repository 300 m
deep in dry tuff for civilian and defense nuclear waste (total capacity ~7000 t IHM), (iii) the
Swedish KBS project for disposal in granite; the Stripa mine has been studied for more than a
decade by an international team and a new underground laboratory has been built at Äspö; other
hard rock laboratories are in operation in Canada, Finland and Switzerland, and (iv) a
laboratory at 250 m depth in clay at Mol, Belgium, has been in operation since 1983, etc. These
sites are all located in geologic formations, which have been unaltered for more than 100
million years. From the tectonic plate theory continual tectonic stability is expected for at least
the next 10 million years. Although ice ages may alter surface conditions at northerly located
sites, they are expected to have a negligible effect on repositories at great depth (>500 m). In
2001 the Finnish parliament gave its permission to begin construction of a KBS-like final
repository for spent fuel in granite. It will be the first of its kind in the world.
21.14. Beneficial utilization of nuclear wastes
The amount of spent fuel generated during 1994 amounts to ~8000 t. It contains large and
potentially valuable sources of metals and radioactive nuclides. Though today considered a
liability it may in the future become a needed asset. Since the extraction and utilization of some
of the fission products or actinides will probably not be economic after the waste has been
vitrified and placed in permanent geologic storage, the nuclear fuel reprocessing scheme should
therefore be designed for byproduct extraction.
Presently the most interesting products in the waste are the platinum group metals (due to their
metal values) and 90Sr, 137Cs, 85Kr, 237Np/238Pu, and 241Am (due to their radiation properties,
cf. §§7.11, 9.5, etc).
The waste contains considerable amounts of Ru, Rh, and Pd, all metals in scarce abundance
on earth. These elements are used as catalysts in the chemical industry, for catalytic exhaust
cleaning in cars, and as corrosion resistant materials. The United States demand exceeds the
domestic production by about a factor of 100. The United States would be independent of
import from year 2000 if these elements were recovered from the yearly generated spent fuel.
This is particularly true if technetium is recovered, since it can often replace platinum. Some
of the recovered elements would be radioactive, but the activities would be small enough to
make the elements easy to handle.
Beta radiation from 85Kr on phosphors causes visible light. Radiokrypton light sources have
widespread applications where reliable lights are required as, for example, at airports,
railroads, hospitals, etc., or where sources of electricity could cause dangerous explosions, as
in coal mines, natural gas plants, etc. However, fission product krypton contains only 4% of
85
Kr, which makes it unsuitable for high intensity lightning applications. 85Kr must therefore
be enriched about a factor of 10, which presently can be done e.g. by thermal diffusion.
By the year 2000, over 100 MW heat will be produced by radiostrontium. Strontium fueled
thermoelectric generators are used in several countries to power unmanned weather data
acquisition systems, lighthouses, and other navigation aids, etc. Their reliability
640
surpasses any other remote power source. The current thermoelectric generators have a
thermal-to-electrical efficiency of about 5%, while recently developed thermal-to-mechanical
systems show efficiencies of 25!30%. The use of such systems could be expanded vastly by
increased recovery of suitable fission products and actinides.
Food sterilization by radiation is potentially of global importance (§18.7). Though this
presently is done by 60Co or accelerator radiation, the advent of large quantities of radiocesium
recovered from nuclear wastes may have a considerable positive impact on the economics and
scale of food irradiation. An almost equally important use of 137Cs would be for sewage sludge
treatment, which may become increasingly important as the requirements for sterilization and
secondary treatment of the sludge increases. The thermoradiation of sludge may also make it
useful as a sterile fertilizer.
The low penetrating radiation, long half-life, and high power density of 238Pu makes it ideal
for special purpose power supplies (§6.9.3). The main present use is in space research, and
~30 kWth power sources have been launched into space, 238Pu has been used in heart
pacemakers and is still a candidate as a power source for completely artificial hearts. Production
of 238Pu requires the isolation of 237Np, which is then irradiated and reprocessed to produce
pure 238Pu.
The demand for 241Am is larger than present production capacity because of its use in logging
oil wells, in smoke detectors, and for various gauging and metering devices. However, the
potential source is large: the annual world production is ~4.5 t in 1994 (after 10 years cooling,
see §21.32.2) and the accumulated civilian nuclear wastes in the US contain >20 t of
americium.
If all the mentioned waste products are recovered, this would mean (a) that the waste is turned
into an essential asset with benefits in food production, health, and safety, and (b) that the
hazard of the remaining waste would be much lower, considerably simplifying final waste
storage.
21.15. Exercises
21.1. In a BWR the minimum and maximum heat fluxes at the fuel rod surface are 0.46 and 1.08 MW m!2 at a coolant
temperature of 283EC. The rods have an outer diameter of 12.7 mm with a cladding of 0.8 mm thick Zircaloy-4; assume
a negligible fuel cladding gap and neglect the temperature drop between coolant and cladding and across the fuel-cladding
gap. Assume that data in Table 21.1 are valid at all temperatures and kUO2 = 4 W/mEC. What are the highest and lowest
fuel-center temperatures?
21.2. Using the data for the Würgassen reactor and the thermal neutron capture cross-section (Table 19.4) it can be
calculated how many kg Pu should be formed per t U at a burn-up of 27 500 MWd/t. (a) Make this calculation assuming
that plutonium disappears only through fission in 239Pu. (b) According to Table 21.2 each t U from a PWR contains 8.69
kg Pu; why is your result much lower?
21.3. It is desired that 98% of all 233Th formed by neutron capture in 232Th decays to 233U. How long a time must
elapse between end of irradiation and start of reprocessing?
21.4. In a 233U fueled reactor, some 233U is converted into 235U. Calculate the amount of 235U formed in 1 t 233U from
neutron capture in 233U and 234U (Fn,( 97 b) for a fluence of 1025 n m!2 (a) assuming no consumption of 235U formed,
(b) taking 235U fission and capture into account.
21.5. Explain why 228Th and 232U are a nuisance in the thorium-uranium fuel cycle.
21.6. A reactor starting with 3% 235U produces 6000 MWd energy/t U fuel each year. Neglecting fission in 238U, (a)
how much fission products have been produced after 5 years? (b) What is the 235U concentration if plutonium fission also
is taken into account?
641
21.7. In the example above, 1 t U as fuel elements is removed from the reactor after 2 years. Using Fig. 21.7, (a) what
is the total radioactivity from the fission products after 1 y cooling time? (b) Which FP elements are the most radioactive
ones at this time?
21.8. Calculate the decontamination factor required for (a) fission product activity, and (b) for gadolinium in
commercial plutonium nitrate produced from PWR fuel (Tables 21.2 and 21.7) at tcool 10 y.
21.9. 0.0015 Ci 239Pu is released annually from a reprocessing plant. What will be the corresponding release of 238Pu
and 240Pu for typical isotopic plutonium composition of LWR fuel?
21.10. Calculate the natural radiotoxicity value Inw of 1 km3 of land (density 2 600 kg m!3) containing 3 weight ppm
238
U with daughter products. Only 226Ra has to be considered.
21.11. A tank contains 100 m3 5 y old HLLW. Analyses show that a 1 ml sample contains 1.09 GBq of 90Sr, which
is the only Sr activity. (a) Calculate the heat production for a waste of composition in Table 21.9 left column. (b) How
many 500 kg glass cylinders would be needed (assume the glass contains 10% FPs) to contain all the solidified waste?
(c) How many 1000 MWe PWR reactor years does this waste correspond to?
21.12. Ru, Rh, and Pd are recovered from the waste from a 10 GWe program. What will the annual amounts and
specific radioactivities be at tcool = 10 y for each of them?
21.16. Literature
J. PRAWITZ and J. RYDBERG, Composition of products formed by thermal neutron fission of 235U, Acta Chem. Scand.
12 (1958) 369, 377.
R. STEPHENSON, Introduction to Nuclear Engineering, McGraw-Hill, 1958.
J. FLAGG (Ed.), Chemical Processing of Reactor Fuels, Academic Press, 1961.
J. T. LONG, Engineering for Nuclear Fuel Reprocessing, Gordon & Breach, 1967.
K. J. SCHNEIDER and A. M. PLATT (Eds.), High-level Radioactive Waste Management Alternatives, Battelle Pacific
Northwest Laboratories, Richland, Wash. 99352, Report BNWL-1900, NTIS, Springfield, 1974.
S. AHRLAND. J. O. LILJENZIN and J. RYDBERG, (Actinide) Solution Chemistry, in Comprehensive Inorganic Chemistry,
Vol. 5, Pergamon Press, Oxford, 1975.
Convention on the Prevention of Marine Pollution by Dumping of Wastes and other Matter ! "The Definition Required
by Annex I, paragraph 6 to the Convention and the Recommendations Required by Annex II, Section D", IAEA,
INFCIRC/205/Add. 1, 10 January 1975.
IAEA, Transuranium Nuclides in the Environment, Vienna, 1976.
IAEA, Management of Radioactive Wastes from the Nuclear Fuel Cycle, Vienna, 1976.
ERDA, Alternatives for Managing Wastes from Reactors and Postfission Operations in the LWR Fuel Cycle,
ERDA-76-43. NTIS, Springfield, 1976.
IAEA, Regional Nuclear Fuel Centres, Vienna, 1977.
NEA-OECD, Objectives, Concepts and Strategies for the Management of Radioactive Waste Arising from Nuclear Power
Programmes, OECD, 1977.
H. A. C. MCKAY and M. G. SOWERBY, The Separation and Recycling of Actinides, EEC, EUR 5801e, 1977.
KBS Project, Nuclear Fuel Cycle Back-End, 1978, Kärnbränslesäkerhet, Fack, 10240 Stockholm.
R. G. WYMER and B. L. VONDRA, Technology of the Light Water Reactor Fuel Cycle, CRC-Press, 1980.
M. BENEDICT, T. H. PIGFORD and H. W. LEVI, Nuclear Chemical Engineering, 2nd Ed., McGraw-Hill, 1981.
C. MUSIKAS and W. SCHULZ, Ch. 11 in J. Rydberg, C. Musikas and G. Choppin, Principles and Practices of Solvent
Extraction, Marcel Dekker, 1992.
SKB, Swedish Nuclear Fuel and Waste Management Company, Annual and other reports, Stockholm, Sweden.
Sources and Effects of Ionizing Radiation (UNSCEAR), United Nations, New York, 1993.
IAEA Bulletin 36, No 1, 1994. (Radiation Technologies)

The Nuclear Fuel Cycle

Transcription

Similar documents

Window Sticker

Window Sticker - FordDirect.com

Hiller / Bell 47 Turbine Helicopter Technical Specifications and

BioBor JF Spec Sheet 2

BioBor JF Spec Sheet 2

John Sheehan - State Ag and Rural Leaders

2014 valkyrie - Honda Powersports

Bleeding Air from Diesel Fuel Lines and Filters

suzuki - Fuelpak

Member Discount Fuel Program - Terre Haute Chamber of Commerce